Token | silly |
Token ID | 14397 |
Feature activation | -9.47e-01 |
Pos logit contributions | |
brand | +0.211 |
truly | +0.187 |
� | +0.175 |
sky | +0.172 |
breathtaking | +0.172 |
Neg logit contributions | |
ried | -0.364 |
ases | -0.342 |
ck | -0.338 |
aunts | -0.318 |
race | -0.313 |
Token | joke |
Token ID | 9707 |
Feature activation | -8.50e-01 |
Pos logit contributions | |
brand | +0.029 |
sky | +0.027 |
truly | +0.026 |
� | +0.026 |
Not | +0.025 |
Neg logit contributions | |
ried | -0.032 |
ck | -0.029 |
ases | -0.029 |
aunts | -0.027 |
beh | -0.026 |
Token | . |
Token ID | 13 |
Feature activation | +2.57e+00 |
Pos logit contributions | |
brand | +0.031 |
truly | +0.029 |
sky | +0.028 |
� | +0.028 |
Not | +0.028 |
Neg logit contributions | |
ried | -0.023 |
ck | -0.021 |
ases | -0.019 |
aunts | -0.019 |
ucks | -0.018 |
Token |
|
Token ID | 198 |
Feature activation | -3.70e+00 |
Pos logit contributions | |
ried | +0.093 |
ck | +0.087 |
ases | +0.085 |
aunts | +0.079 |
beh | +0.076 |
Neg logit contributions | |
brand | -0.069 |
� | -0.063 |
truly | -0.062 |
sky | -0.060 |
Not | -0.060 |
Token |
|
Token ID | 198 |
Feature activation | -3.91e+00 |
Pos logit contributions | |
brand | +0.080 |
� | +0.070 |
truly | +0.069 |
sky | +0.066 |
Imagine | +0.065 |
Neg logit contributions | |
ried | -0.154 |
ck | -0.145 |
ases | -0.142 |
aunts | -0.134 |
race | -0.130 |
Token | Tam |
Token ID | 42061 |
Feature activation | +1.56e+01 |
Pos logit contributions | |
brand | +0.086 |
truly | +0.077 |
� | +0.077 |
Imagine | +0.074 |
Not | +0.074 |
Neg logit contributions | |
ried | -0.161 |
ck | -0.149 |
ases | -0.147 |
aunts | -0.138 |
ped | -0.135 |
Token | my |
Token ID | 1820 |
Feature activation | +4.52e+00 |
Pos logit contributions | |
ried | +0.479 |
ases | +0.460 |
ck | +0.453 |
ucks | +0.435 |
beh | +0.432 |
Neg logit contributions | |
brand | -0.444 |
sky | -0.444 |
truly | -0.438 |
Imagine | -0.429 |
fact | -0.417 |
Token | was |
Token ID | 373 |
Feature activation | -3.65e+00 |
Pos logit contributions | |
ried | +0.164 |
ases | +0.162 |
ck | +0.153 |
ucks | +0.148 |
aunts | +0.145 |
Neg logit contributions | |
brand | -0.112 |
truly | -0.108 |
sky | -0.104 |
Imagine | -0.102 |
Not | -0.101 |
Token | very |
Token ID | 845 |
Feature activation | -1.00e+01 |
Pos logit contributions | |
brand | +0.074 |
truly | +0.064 |
� | +0.064 |
sky | +0.060 |
Not | +0.059 |
Neg logit contributions | |
ried | -0.157 |
ck | -0.148 |
ases | -0.145 |
aunts | -0.137 |
race | -0.133 |
Token | pleased |
Token ID | 10607 |
Feature activation | -4.04e+00 |
Pos logit contributions | |
brand | +0.219 |
truly | +0.215 |
Not | +0.198 |
much | +0.184 |
Imagine | +0.184 |
Neg logit contributions | |
ried | -0.414 |
ases | -0.379 |
ck | -0.374 |
ucks | -0.363 |
race | -0.357 |
Token | , |
Token ID | 11 |
Feature activation | -2.64e+00 |
Pos logit contributions | |
brand | +0.152 |
truly | +0.141 |
� | +0.140 |
sky | +0.137 |
Not | +0.136 |
Neg logit contributions | |
ried | -0.106 |
ck | -0.096 |
ases | -0.092 |
aunts | -0.083 |
race | -0.079 |
Token | the |
Token ID | 262 |
Feature activation | -1.31e-01 |
Pos logit contributions | |
ried | +0.084 |
ck | +0.076 |
ases | +0.071 |
aunts | +0.063 |
beh | +0.061 |
Neg logit contributions | |
brand | -0.161 |
truly | -0.150 |
� | -0.150 |
sky | -0.149 |
Not | -0.145 |
Token | hive |
Token ID | 35881 |
Feature activation | -3.99e+00 |
Pos logit contributions | |
brand | +0.004 |
� | +0.004 |
Not | +0.004 |
Imagine | +0.004 |
Even | +0.004 |
Neg logit contributions | |
ried | -0.003 |
ck | -0.003 |
leave | -0.003 |
ie | -0.003 |
ases | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | +2.44e+00 |
Pos logit contributions | |
brand | +0.171 |
truly | +0.160 |
� | +0.159 |
sky | +0.158 |
breathtaking | +0.154 |
Neg logit contributions | |
ried | -0.081 |
ck | -0.073 |
ases | -0.070 |
aunts | -0.062 |
beh | -0.058 |
Token |
|
Token ID | 198 |
Feature activation | -3.73e+00 |
Pos logit contributions | |
ried | +0.066 |
race | +0.061 |
ases | +0.061 |
aunts | +0.059 |
Kim | +0.058 |
Neg logit contributions | |
brand | -0.085 |
truly | -0.076 |
breathtaking | -0.074 |
fact | -0.071 |
� | -0.071 |
Token |
|
Token ID | 198 |
Feature activation | -3.79e+00 |
Pos logit contributions | |
brand | +0.106 |
� | +0.096 |
truly | +0.094 |
sky | +0.091 |
breathtaking | +0.090 |
Neg logit contributions | |
ried | -0.128 |
ck | -0.122 |
ases | -0.117 |
aunts | -0.110 |
race | -0.107 |
Token | Tam |
Token ID | 42061 |
Feature activation | +1.55e+01 |
Pos logit contributions | |
brand | +0.110 |
� | +0.098 |
truly | +0.097 |
sky | +0.094 |
breathtaking | +0.093 |
Neg logit contributions | |
ried | -0.130 |
ck | -0.122 |
ases | -0.118 |
aunts | -0.111 |
race | -0.109 |
Token | my |
Token ID | 1820 |
Feature activation | +4.51e+00 |
Pos logit contributions | |
ried | +0.451 |
ases | +0.441 |
ck | +0.432 |
steer | +0.406 |
beh | +0.396 |
Neg logit contributions | |
brand | -0.486 |
sky | -0.468 |
truly | -0.465 |
Imagine | -0.454 |
fact | -0.444 |
Token | went |
Token ID | 1816 |
Feature activation | +8.46e+00 |
Pos logit contributions | |
ried | +0.161 |
ck | +0.150 |
ases | +0.146 |
aunts | +0.137 |
race | +0.132 |
Neg logit contributions | |
brand | -0.124 |
� | -0.112 |
truly | -0.111 |
sky | -0.106 |
breathtaking | -0.106 |
Token | right |
Token ID | 826 |
Feature activation | +4.57e+00 |
Pos logit contributions | |
ried | +0.226 |
ck | +0.225 |
ases | +0.224 |
aunts | +0.202 |
beh | +0.199 |
Neg logit contributions | |
brand | -0.304 |
truly | -0.277 |
sky | -0.277 |
Once | -0.261 |
� | -0.261 |
Token | up |
Token ID | 510 |
Feature activation | +2.36e+00 |
Pos logit contributions | |
ried | +0.096 |
ck | +0.092 |
ases | +0.091 |
aunts | +0.081 |
beh | +0.080 |
Neg logit contributions | |
brand | -0.186 |
sky | -0.174 |
truly | -0.172 |
breathtaking | -0.168 |
� | -0.164 |
Token | to |
Token ID | 284 |
Feature activation | +6.50e+00 |
Pos logit contributions | |
ried | +0.061 |
ck | +0.056 |
ases | +0.054 |
aunts | +0.049 |
beh | +0.047 |
Neg logit contributions | |
brand | -0.088 |
� | -0.082 |
truly | -0.082 |
sky | -0.081 |
Not | -0.079 |
Token | the |
Token ID | 262 |
Feature activation | +6.75e-01 |
Pos logit contributions | |
ried | +0.024 |
ck | +0.021 |
ases | +0.021 |
aunts | +0.020 |
race | +0.019 |
Neg logit contributions | |
brand | -0.027 |
truly | -0.025 |
sky | -0.024 |
Not | -0.024 |
breathtaking | -0.024 |
Token | toys |
Token ID | 14958 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
ried | +0.022 |
ck | +0.021 |
ases | +0.020 |
aunts | +0.019 |
race | +0.018 |
Neg logit contributions | |
brand | -0.020 |
� | -0.018 |
truly | -0.018 |
Imagine | -0.017 |
Not | -0.017 |
Token | ." |
Token ID | 526 |
Feature activation | +2.35e+00 |
Pos logit contributions | |
brand | +0.055 |
� | +0.051 |
truly | +0.050 |
sky | +0.049 |
Not | +0.049 |
Neg logit contributions | |
ried | -0.044 |
ck | -0.040 |
ases | -0.039 |
aunts | -0.036 |
race | -0.034 |
Token |
|
Token ID | 198 |
Feature activation | -3.39e+00 |
Pos logit contributions | |
ried | +0.124 |
ck | +0.117 |
ases | +0.112 |
beh | +0.111 |
rying | +0.109 |
Neg logit contributions | |
Not | -0.029 |
� | -0.028 |
sky | -0.026 |
brand | -0.024 |
Cookies | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | -3.59e+00 |
Pos logit contributions | |
brand | +0.034 |
Finally | +0.033 |
Not | +0.031 |
sky | +0.028 |
� | +0.028 |
Neg logit contributions | |
ried | -0.184 |
ases | -0.163 |
beh | -0.162 |
ck | -0.160 |
rying | -0.157 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.55e+01 |
Pos logit contributions | |
Finally | +0.090 |
Eventually | +0.080 |
Imagine | +0.080 |
Not | +0.080 |
� | +0.076 |
Neg logit contributions | |
ried | -0.147 |
ases | -0.126 |
ped | -0.124 |
leave | -0.119 |
ck | -0.119 |
Token | Lee |
Token ID | 5741 |
Feature activation | +7.61e+00 |
Pos logit contributions | |
ck | +0.447 |
ried | +0.400 |
steer | +0.377 |
ases | +0.363 |
aunts | +0.344 |
Neg logit contributions | |
brand | -0.512 |
� | -0.497 |
sky | -0.490 |
Imagine | -0.481 |
Not | -0.477 |
Token | looked |
Token ID | 3114 |
Feature activation | -5.12e+00 |
Pos logit contributions | |
ried | +0.255 |
ck | +0.232 |
ases | +0.231 |
beh | +0.216 |
aunts | +0.215 |
Neg logit contributions | |
brand | -0.225 |
truly | -0.207 |
� | -0.206 |
sky | -0.205 |
Not | -0.202 |
Token | at |
Token ID | 379 |
Feature activation | +6.36e+00 |
Pos logit contributions | |
brand | +0.211 |
truly | +0.190 |
sky | +0.185 |
breathtaking | +0.183 |
Imagine | +0.180 |
Neg logit contributions | |
ried | -0.122 |
ck | -0.109 |
ases | -0.107 |
aunts | -0.097 |
race | -0.096 |
Token | them |
Token ID | 606 |
Feature activation | +3.10e-01 |
Pos logit contributions | |
ried | +0.163 |
ck | +0.137 |
beh | +0.133 |
ases | +0.129 |
aunts | +0.123 |
Neg logit contributions | |
brand | -0.245 |
sky | -0.235 |
� | -0.229 |
Not | -0.227 |
fact | -0.226 |
Token | . |
Token ID | 13 |
Feature activation | +2.46e+00 |
Pos logit contributions | |
ried | +0.009 |
ases | +0.008 |
ck | +0.008 |
aunts | +0.007 |
beh | +0.007 |
Neg logit contributions | |
brand | -0.011 |
� | -0.010 |
truly | -0.010 |
sky | -0.010 |
Not | -0.010 |
Token | mistake |
Token ID | 7457 |
Feature activation | +6.91e-01 |
Pos logit contributions | |
� | +0.106 |
brand | +0.106 |
Not | +0.098 |
Imagine | +0.098 |
truly | +0.096 |
Neg logit contributions | |
ck | -0.088 |
ried | -0.088 |
ases | -0.080 |
ies | -0.078 |
aunts | -0.077 |
Token | . |
Token ID | 13 |
Feature activation | +3.23e+00 |
Pos logit contributions | |
ried | +0.017 |
ck | +0.016 |
ases | +0.015 |
aunts | +0.014 |
beh | +0.013 |
Neg logit contributions | |
brand | -0.026 |
truly | -0.025 |
sky | -0.025 |
Not | -0.024 |
� | -0.024 |
Token |
|
Token ID | 198 |
Feature activation | -3.21e+00 |
Pos logit contributions | |
ried | +0.098 |
ases | +0.092 |
ck | +0.089 |
aunts | +0.085 |
ucks | +0.083 |
Neg logit contributions | |
brand | -0.103 |
truly | -0.093 |
� | -0.090 |
breathtaking | -0.089 |
Imagine | -0.087 |
Token |
|
Token ID | 198 |
Feature activation | -3.26e+00 |
Pos logit contributions | |
brand | +0.069 |
� | +0.061 |
truly | +0.060 |
sky | +0.057 |
Imagine | +0.056 |
Neg logit contributions | |
ried | -0.133 |
ck | -0.126 |
ases | -0.123 |
aunts | -0.116 |
race | -0.113 |
Token | " |
Token ID | 1 |
Feature activation | +6.31e-01 |
Pos logit contributions | |
brand | +0.103 |
� | +0.094 |
truly | +0.094 |
sky | +0.091 |
Imagine | +0.091 |
Neg logit contributions | |
ried | -0.104 |
ck | -0.095 |
ases | -0.093 |
aunts | -0.085 |
beh | -0.083 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.55e+01 |
Pos logit contributions | |
ried | +0.021 |
ck | +0.020 |
ases | +0.019 |
aunts | +0.018 |
beh | +0.017 |
Neg logit contributions | |
brand | -0.019 |
� | -0.017 |
truly | -0.017 |
breathtaking | -0.016 |
Not | -0.016 |
Token | , |
Token ID | 11 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
ck | +0.432 |
ried | +0.422 |
ases | +0.364 |
aunts | +0.358 |
steer | +0.356 |
Neg logit contributions | |
brand | -0.512 |
Not | -0.490 |
sky | -0.486 |
truly | -0.481 |
� | -0.472 |
Token | we |
Token ID | 356 |
Feature activation | +3.76e+00 |
Pos logit contributions | |
brand | +0.035 |
� | +0.032 |
truly | +0.032 |
sky | +0.031 |
Imagine | +0.031 |
Neg logit contributions | |
ried | -0.032 |
ck | -0.029 |
ases | -0.029 |
aunts | -0.026 |
race | -0.025 |
Token | have |
Token ID | 423 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
ried | +0.140 |
ck | +0.131 |
ases | +0.128 |
aunts | +0.120 |
race | +0.116 |
Neg logit contributions | |
brand | -0.098 |
� | -0.087 |
truly | -0.087 |
sky | -0.083 |
Not | -0.082 |
Token | your |
Token ID | 534 |
Feature activation | +3.02e+00 |
Pos logit contributions | |
ried | +0.034 |
ck | +0.031 |
ases | +0.030 |
aunts | +0.028 |
race | +0.026 |
Neg logit contributions | |
brand | -0.044 |
truly | -0.040 |
� | -0.040 |
sky | -0.039 |
Not | -0.039 |
Token | photo |
Token ID | 4590 |
Feature activation | -5.78e+00 |
Pos logit contributions | |
ried | +0.083 |
ases | +0.078 |
ck | +0.078 |
ies | +0.070 |
leave | +0.069 |
Neg logit contributions | |
brand | -0.101 |
Imagine | -0.095 |
� | -0.095 |
Finally | -0.093 |
breathtaking | -0.093 |
Token | protect |
Token ID | 1805 |
Feature activation | +5.39e+00 |
Pos logit contributions | |
ried | +0.191 |
ases | +0.166 |
ucks | +0.156 |
ck | +0.153 |
aunts | +0.152 |
Neg logit contributions | |
truly | -0.215 |
brand | -0.211 |
� | -0.205 |
Not | -0.201 |
sky | -0.195 |
Token | it |
Token ID | 340 |
Feature activation | -8.31e+00 |
Pos logit contributions | |
ried | +0.169 |
ases | +0.150 |
ck | +0.145 |
aunts | +0.144 |
ucks | +0.139 |
Neg logit contributions | |
brand | -0.162 |
truly | -0.159 |
Not | -0.151 |
sky | -0.150 |
lot | -0.145 |
Token | . |
Token ID | 13 |
Feature activation | +2.43e+00 |
Pos logit contributions | |
brand | +0.328 |
truly | +0.312 |
� | +0.294 |
sky | +0.294 |
Not | +0.293 |
Neg logit contributions | |
ried | -0.197 |
ases | -0.174 |
ck | -0.171 |
aunts | -0.164 |
ucks | -0.157 |
Token |
|
Token ID | 198 |
Feature activation | -3.80e+00 |
Pos logit contributions | |
ried | +0.093 |
ck | +0.086 |
ases | +0.085 |
aunts | +0.081 |
race | +0.079 |
Neg logit contributions | |
brand | -0.060 |
truly | -0.053 |
� | -0.052 |
breathtaking | -0.051 |
Imagine | -0.049 |
Token |
|
Token ID | 198 |
Feature activation | -3.83e+00 |
Pos logit contributions | |
brand | +0.080 |
� | +0.070 |
truly | +0.069 |
sky | +0.065 |
Not | +0.065 |
Neg logit contributions | |
ried | -0.160 |
ck | -0.150 |
ases | -0.148 |
aunts | -0.139 |
race | -0.135 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.54e+01 |
Pos logit contributions | |
brand | +0.135 |
� | +0.124 |
truly | +0.124 |
sky | +0.120 |
Imagine | +0.120 |
Neg logit contributions | |
ried | -0.107 |
ck | -0.097 |
ases | -0.095 |
aunts | -0.086 |
race | -0.082 |
Token | y |
Token ID | 88 |
Feature activation | +8.59e-01 |
Pos logit contributions | |
ck | +0.476 |
steer | +0.422 |
ried | +0.418 |
ases | +0.412 |
aunts | +0.378 |
Neg logit contributions | |
brand | -0.485 |
Imagine | -0.455 |
truly | -0.449 |
� | -0.446 |
sky | -0.441 |
Token | thought |
Token ID | 1807 |
Feature activation | -2.21e+00 |
Pos logit contributions | |
ried | +0.029 |
ases | +0.028 |
ck | +0.027 |
aunts | +0.025 |
ucks | +0.025 |
Neg logit contributions | |
brand | -0.024 |
truly | -0.023 |
sky | -0.022 |
� | -0.021 |
Imagine | -0.021 |
Token | about |
Token ID | 546 |
Feature activation | +6.37e-01 |
Pos logit contributions | |
brand | +0.059 |
truly | +0.053 |
� | +0.052 |
sky | +0.050 |
Not | +0.050 |
Neg logit contributions | |
ried | -0.081 |
ck | -0.075 |
ases | -0.074 |
aunts | -0.069 |
race | -0.067 |
Token | this |
Token ID | 428 |
Feature activation | -5.32e+00 |
Pos logit contributions | |
ried | +0.020 |
ck | +0.019 |
ases | +0.018 |
aunts | +0.017 |
race | +0.016 |
Neg logit contributions | |
brand | -0.020 |
� | -0.018 |
truly | -0.017 |
sky | -0.017 |
Imagine | -0.017 |
Token | for |
Token ID | 329 |
Feature activation | +5.34e+00 |
Pos logit contributions | |
� | +0.170 |
brand | +0.170 |
Imagine | +0.161 |
Finally | +0.156 |
Not | +0.153 |
Neg logit contributions | |
ried | -0.149 |
ck | -0.149 |
ies | -0.131 |
ie | -0.130 |
leave | -0.128 |
Token | loud |
Token ID | 7812 |
Feature activation | +2.32e+00 |
Pos logit contributions | |
ried | +0.039 |
ck | +0.031 |
ases | +0.030 |
aunts | +0.029 |
ped | +0.029 |
Neg logit contributions | |
brand | -0.043 |
truly | -0.038 |
sky | -0.037 |
Not | -0.037 |
breathtaking | -0.036 |
Token | noises |
Token ID | 26782 |
Feature activation | -4.17e+00 |
Pos logit contributions | |
ried | +0.072 |
ck | +0.067 |
ases | +0.065 |
aunts | +0.060 |
race | +0.058 |
Neg logit contributions | |
brand | -0.074 |
� | -0.068 |
truly | -0.067 |
Imagine | -0.064 |
sky | -0.064 |
Token | . |
Token ID | 13 |
Feature activation | +2.66e+00 |
Pos logit contributions | |
brand | +0.164 |
truly | +0.153 |
sky | +0.152 |
� | +0.149 |
Not | +0.148 |
Neg logit contributions | |
ried | -0.101 |
ck | -0.086 |
ases | -0.084 |
aunts | -0.078 |
beh | -0.074 |
Token |
|
Token ID | 198 |
Feature activation | -3.49e+00 |
Pos logit contributions | |
ried | +0.079 |
ases | +0.075 |
ck | +0.073 |
aunts | +0.070 |
ucks | +0.067 |
Neg logit contributions | |
brand | -0.086 |
truly | -0.077 |
� | -0.075 |
breathtaking | -0.075 |
Imagine | -0.072 |
Token |
|
Token ID | 198 |
Feature activation | -3.61e+00 |
Pos logit contributions | |
brand | +0.072 |
� | +0.062 |
truly | +0.061 |
sky | +0.059 |
Not | +0.059 |
Neg logit contributions | |
ried | -0.151 |
ck | -0.139 |
ases | -0.138 |
aunts | -0.129 |
beh | -0.128 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.54e+01 |
Pos logit contributions | |
brand | +0.094 |
Finally | +0.093 |
� | +0.089 |
Imagine | +0.088 |
Not | +0.088 |
Neg logit contributions | |
ried | -0.135 |
ases | -0.115 |
ck | -0.113 |
ped | -0.110 |
beh | -0.107 |
Token | Lee |
Token ID | 5741 |
Feature activation | +7.41e+00 |
Pos logit contributions | |
ck | +0.443 |
ried | +0.380 |
steer | +0.362 |
aunts | +0.332 |
beh | +0.331 |
Neg logit contributions | |
brand | -0.502 |
sky | -0.495 |
� | -0.490 |
Imagine | -0.479 |
Not | -0.472 |
Token | says |
Token ID | 1139 |
Feature activation | +5.72e+00 |
Pos logit contributions | |
ried | +0.262 |
ck | +0.237 |
ases | +0.227 |
beh | +0.222 |
aunts | +0.218 |
Neg logit contributions | |
brand | -0.208 |
sky | -0.191 |
truly | -0.188 |
� | -0.187 |
Not | -0.186 |
Token | , |
Token ID | 11 |
Feature activation | -1.87e+00 |
Pos logit contributions | |
ried | +0.200 |
ck | +0.185 |
ases | +0.180 |
aunts | +0.169 |
beh | +0.162 |
Neg logit contributions | |
brand | -0.167 |
� | -0.151 |
truly | -0.150 |
sky | -0.146 |
Not | -0.145 |
Token | " |
Token ID | 366 |
Feature activation | -3.13e+00 |
Pos logit contributions | |
brand | +0.046 |
truly | +0.041 |
breathtaking | +0.040 |
� | +0.040 |
Imagine | +0.038 |
Neg logit contributions | |
ried | -0.070 |
ck | -0.064 |
ases | -0.063 |
aunts | -0.060 |
race | -0.058 |
Token | Thank |
Token ID | 10449 |
Feature activation | +4.36e+00 |
Pos logit contributions | |
brand | +0.093 |
� | +0.082 |
truly | +0.082 |
breathtaking | +0.079 |
sky | +0.077 |
Neg logit contributions | |
ried | -0.107 |
ck | -0.100 |
ases | -0.098 |
aunts | -0.092 |
race | -0.089 |
Token | their |
Token ID | 511 |
Feature activation | +2.45e+00 |
Pos logit contributions | |
ried | +0.150 |
ck | +0.139 |
ases | +0.136 |
aunts | +0.127 |
beh | +0.122 |
Neg logit contributions | |
brand | -0.126 |
truly | -0.113 |
� | -0.112 |
sky | -0.109 |
Not | -0.108 |
Token | dreams |
Token ID | 10625 |
Feature activation | -6.02e+00 |
Pos logit contributions | |
ried | +0.065 |
ck | +0.060 |
ases | +0.058 |
aunts | +0.052 |
race | +0.049 |
Neg logit contributions | |
brand | -0.088 |
� | -0.083 |
truly | -0.081 |
Imagine | -0.080 |
breathtaking | -0.079 |
Token | . |
Token ID | 13 |
Feature activation | +2.68e+00 |
Pos logit contributions | |
brand | +0.248 |
� | +0.231 |
truly | +0.231 |
sky | +0.226 |
Not | +0.223 |
Neg logit contributions | |
ried | -0.131 |
ck | -0.117 |
ases | -0.112 |
aunts | -0.099 |
race | -0.093 |
Token |
|
Token ID | 198 |
Feature activation | -3.67e+00 |
Pos logit contributions | |
ried | +0.084 |
ck | +0.078 |
ases | +0.077 |
aunts | +0.071 |
ucks | +0.069 |
Neg logit contributions | |
brand | -0.083 |
truly | -0.076 |
� | -0.075 |
breathtaking | -0.073 |
Imagine | -0.073 |
Token |
|
Token ID | 198 |
Feature activation | -3.77e+00 |
Pos logit contributions | |
brand | +0.074 |
� | +0.064 |
truly | +0.063 |
sky | +0.060 |
Not | +0.060 |
Neg logit contributions | |
ried | -0.159 |
ck | -0.148 |
ases | -0.146 |
aunts | -0.138 |
beh | -0.136 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.53e+01 |
Pos logit contributions | |
brand | +0.110 |
� | +0.103 |
Finally | +0.102 |
Imagine | +0.102 |
truly | +0.102 |
Neg logit contributions | |
ried | -0.127 |
ck | -0.113 |
ases | -0.112 |
ped | -0.103 |
aunts | -0.102 |
Token | Lee |
Token ID | 5741 |
Feature activation | +7.47e+00 |
Pos logit contributions | |
ck | +0.409 |
ried | +0.389 |
ases | +0.348 |
steer | +0.342 |
aunts | +0.331 |
Neg logit contributions | |
brand | -0.537 |
� | -0.508 |
sky | -0.505 |
truly | -0.497 |
Imagine | -0.496 |
Token | nodded |
Token ID | 14464 |
Feature activation | +3.92e+00 |
Pos logit contributions | |
ried | +0.237 |
ases | +0.217 |
ck | +0.214 |
aunts | +0.198 |
beh | +0.196 |
Neg logit contributions | |
brand | -0.236 |
truly | -0.217 |
� | -0.214 |
sky | -0.211 |
Not | -0.207 |
Token | . |
Token ID | 13 |
Feature activation | +2.50e+00 |
Pos logit contributions | |
ried | +0.131 |
ases | +0.123 |
beh | +0.119 |
ck | +0.118 |
race | +0.116 |
Neg logit contributions | |
brand | -0.121 |
sky | -0.109 |
truly | -0.108 |
so | -0.103 |
much | -0.101 |
Token | She |
Token ID | 1375 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
ried | +0.077 |
ases | +0.071 |
ck | +0.069 |
aunts | +0.067 |
ucks | +0.065 |
Neg logit contributions | |
brand | -0.076 |
� | -0.070 |
truly | -0.070 |
breathtaking | -0.068 |
Imagine | -0.065 |
Token | hugged |
Token ID | 41130 |
Feature activation | +9.16e+00 |
Pos logit contributions | |
ried | +0.045 |
ck | +0.041 |
ases | +0.041 |
aunts | +0.038 |
race | +0.037 |
Neg logit contributions | |
brand | -0.042 |
truly | -0.039 |
� | -0.037 |
sky | -0.037 |
Imagine | -0.037 |
Token | that |
Token ID | 326 |
Feature activation | -3.13e+00 |
Pos logit contributions | |
ried | +0.246 |
ck | +0.225 |
ases | +0.218 |
aunts | +0.200 |
race | +0.191 |
Neg logit contributions | |
brand | -0.309 |
� | -0.285 |
truly | -0.283 |
sky | -0.277 |
Not | -0.274 |
Token | bush |
Token ID | 24484 |
Feature activation | -6.16e+00 |
Pos logit contributions | |
brand | +0.104 |
� | +0.098 |
Finally | +0.092 |
Not | +0.089 |
Eventually | +0.089 |
Neg logit contributions | |
ried | -0.092 |
ck | -0.084 |
ases | -0.081 |
ies | -0.078 |
leave | -0.076 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
brand | +0.160 |
� | +0.143 |
truly | +0.142 |
sky | +0.137 |
Imagine | +0.135 |
Neg logit contributions | |
ried | -0.227 |
ck | -0.213 |
ases | -0.208 |
aunts | -0.195 |
race | -0.189 |
Token |
|
Token ID | 198 |
Feature activation | -3.95e+00 |
Pos logit contributions | |
ried | +0.054 |
ases | +0.051 |
race | +0.049 |
ck | +0.049 |
aunts | +0.049 |
Neg logit contributions | |
brand | -0.050 |
breathtaking | -0.047 |
truly | -0.047 |
impossible | -0.044 |
daunting | -0.044 |
Token |
|
Token ID | 198 |
Feature activation | -3.99e+00 |
Pos logit contributions | |
brand | +0.075 |
� | +0.064 |
truly | +0.063 |
sky | +0.060 |
Imagine | +0.060 |
Neg logit contributions | |
ried | -0.179 |
ck | -0.167 |
ases | -0.164 |
aunts | -0.153 |
beh | -0.153 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.53e+01 |
Pos logit contributions | |
brand | +0.153 |
truly | +0.142 |
Imagine | +0.140 |
� | +0.139 |
Finally | +0.139 |
Neg logit contributions | |
ried | -0.103 |
ck | -0.091 |
ases | -0.089 |
aunts | -0.077 |
ped | -0.077 |
Token | y |
Token ID | 88 |
Feature activation | +9.69e-01 |
Pos logit contributions | |
ck | +0.463 |
ried | +0.401 |
steer | +0.391 |
ases | +0.381 |
aunts | +0.358 |
Neg logit contributions | |
brand | -0.500 |
� | -0.476 |
Imagine | -0.461 |
sky | -0.453 |
Not | -0.451 |
Token | rushed |
Token ID | 13999 |
Feature activation | +1.20e+01 |
Pos logit contributions | |
ried | +0.031 |
ases | +0.030 |
ck | +0.029 |
aunts | +0.027 |
race | +0.027 |
Neg logit contributions | |
brand | -0.029 |
truly | -0.027 |
� | -0.027 |
sky | -0.026 |
Not | -0.026 |
Token | over |
Token ID | 625 |
Feature activation | +4.59e+00 |
Pos logit contributions | |
ck | +0.412 |
ases | +0.403 |
ried | +0.384 |
beh | +0.374 |
aunts | +0.364 |
Neg logit contributions | |
brand | -0.334 |
sky | -0.317 |
so | -0.310 |
Once | -0.302 |
� | -0.287 |
Token | and |
Token ID | 290 |
Feature activation | -2.90e-01 |
Pos logit contributions | |
ck | +0.147 |
ried | +0.147 |
ases | +0.139 |
beh | +0.136 |
aunts | +0.132 |
Neg logit contributions | |
brand | -0.138 |
� | -0.132 |
sky | -0.126 |
truly | -0.124 |
Imagine | -0.121 |
Token | sure |
Token ID | 1654 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
brand | +0.010 |
truly | +0.009 |
� | +0.009 |
Imagine | +0.009 |
sky | +0.008 |
Neg logit contributions | |
ases | -0.009 |
ried | -0.008 |
ck | -0.008 |
aunts | -0.007 |
beh | -0.007 |
Token | hug |
Token ID | 16225 |
Feature activation | +5.82e+00 |
Pos logit contributions | |
brand | +0.038 |
� | +0.035 |
truly | +0.035 |
sky | +0.034 |
breathtaking | +0.034 |
Neg logit contributions | |
ried | -0.032 |
ck | -0.029 |
ases | -0.028 |
aunts | -0.026 |
beh | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | +2.29e+00 |
Pos logit contributions | |
ried | +0.157 |
ck | +0.151 |
ases | +0.146 |
aunts | +0.129 |
beh | +0.124 |
Neg logit contributions | |
brand | -0.215 |
truly | -0.200 |
sky | -0.198 |
� | -0.195 |
breathtaking | -0.186 |
Token | |
Token ID | 220 |
Feature activation | -1.18e+01 |
Pos logit contributions | |
ried | +0.072 |
ases | +0.066 |
ck | +0.065 |
aunts | +0.064 |
race | +0.062 |
Neg logit contributions | |
brand | -0.068 |
truly | -0.062 |
� | -0.060 |
breathtaking | -0.059 |
Imagine | -0.058 |
Token |
|
Token ID | 198 |
Feature activation | -3.96e+00 |
Pos logit contributions | |
brand | +0.211 |
� | +0.194 |
Not | +0.180 |
truly | +0.180 |
sky | +0.170 |
Neg logit contributions | |
ried | -0.530 |
ck | -0.508 |
ases | -0.490 |
aunts | -0.463 |
ies | -0.456 |
Token |
|
Token ID | 198 |
Feature activation | -4.01e+00 |
Pos logit contributions | |
brand | +0.091 |
� | +0.079 |
truly | +0.079 |
sky | +0.075 |
breathtaking | +0.074 |
Neg logit contributions | |
ried | -0.159 |
ck | -0.150 |
ases | -0.147 |
aunts | -0.139 |
race | -0.135 |
Token | Tam |
Token ID | 42061 |
Feature activation | +1.53e+01 |
Pos logit contributions | |
brand | +0.107 |
� | +0.095 |
truly | +0.093 |
sky | +0.090 |
breathtaking | +0.089 |
Neg logit contributions | |
ried | -0.146 |
ck | -0.139 |
ases | -0.134 |
aunts | -0.127 |
race | -0.124 |
Token | my |
Token ID | 1820 |
Feature activation | +4.00e+00 |
Pos logit contributions | |
ried | +0.419 |
ases | +0.396 |
ck | +0.381 |
steer | +0.349 |
beh | +0.348 |
Neg logit contributions | |
brand | -0.531 |
truly | -0.503 |
sky | -0.497 |
fact | -0.476 |
Imagine | -0.475 |
Token | was |
Token ID | 373 |
Feature activation | -3.50e+00 |
Pos logit contributions | |
ried | +0.147 |
ck | +0.137 |
ases | +0.136 |
aunts | +0.125 |
beh | +0.123 |
Neg logit contributions | |
brand | -0.106 |
truly | -0.096 |
� | -0.094 |
sky | -0.092 |
Not | -0.089 |
Token | so |
Token ID | 523 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
brand | +0.062 |
� | +0.059 |
Imagine | +0.052 |
sky | +0.052 |
Not | +0.051 |
Neg logit contributions | |
ried | -0.154 |
ck | -0.146 |
ases | -0.140 |
aunts | -0.134 |
leave | -0.132 |
Token | proud |
Token ID | 6613 |
Feature activation | -3.03e+00 |
Pos logit contributions | |
brand | +0.298 |
� | +0.278 |
sky | +0.261 |
Imagine | +0.255 |
truly | +0.253 |
Neg logit contributions | |
ried | -0.410 |
ck | -0.392 |
ases | -0.372 |
aunts | -0.356 |
beh | -0.345 |
Token | of |
Token ID | 286 |
Feature activation | +5.80e+00 |
Pos logit contributions | |
brand | +0.100 |
truly | +0.091 |
� | +0.089 |
sky | +0.089 |
Not | +0.087 |
Neg logit contributions | |
ried | -0.095 |
ck | -0.086 |
ases | -0.084 |
aunts | -0.077 |
beh | -0.074 |
Token | says |
Token ID | 1139 |
Feature activation | +5.46e+00 |
Pos logit contributions | |
ried | +0.004 |
ck | +0.004 |
ases | +0.004 |
aunts | +0.003 |
beh | +0.003 |
Neg logit contributions | |
brand | -0.005 |
Not | -0.004 |
sky | -0.004 |
truly | -0.004 |
� | -0.004 |
Token | proudly |
Token ID | 23017 |
Feature activation | +5.59e+00 |
Pos logit contributions | |
ried | +0.170 |
ases | +0.166 |
ck | +0.159 |
aunts | +0.154 |
ucks | +0.152 |
Neg logit contributions | |
brand | -0.168 |
much | -0.155 |
so | -0.145 |
Not | -0.144 |
truly | -0.141 |
Token | . |
Token ID | 13 |
Feature activation | +2.65e+00 |
Pos logit contributions | |
ried | +0.165 |
ases | +0.146 |
ck | +0.143 |
aunts | +0.137 |
beh | +0.134 |
Neg logit contributions | |
brand | -0.180 |
Not | -0.165 |
sky | -0.165 |
� | -0.163 |
Once | -0.156 |
Token |
|
Token ID | 198 |
Feature activation | -3.72e+00 |
Pos logit contributions | |
ried | +0.121 |
ck | +0.116 |
ases | +0.111 |
beh | +0.106 |
aunts | +0.104 |
Neg logit contributions | |
brand | -0.049 |
� | -0.043 |
Not | -0.042 |
sky | -0.042 |
truly | -0.040 |
Token |
|
Token ID | 198 |
Feature activation | -3.75e+00 |
Pos logit contributions | |
brand | +0.071 |
� | +0.061 |
truly | +0.061 |
sky | +0.058 |
Not | +0.058 |
Neg logit contributions | |
ried | -0.164 |
ck | -0.154 |
ases | -0.151 |
aunts | -0.142 |
beh | -0.141 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.52e+01 |
Pos logit contributions | |
Finally | +0.093 |
Imagine | +0.089 |
brand | +0.089 |
Not | +0.086 |
Eventually | +0.086 |
Neg logit contributions | |
ried | -0.144 |
ck | -0.132 |
ases | -0.127 |
ped | -0.123 |
beh | -0.121 |
Token | Lee |
Token ID | 5741 |
Feature activation | +7.55e+00 |
Pos logit contributions | |
ck | +0.437 |
ried | +0.402 |
steer | +0.348 |
ases | +0.347 |
beh | +0.336 |
Neg logit contributions | |
brand | -0.498 |
sky | -0.490 |
Not | -0.489 |
� | -0.485 |
Imagine | -0.472 |
Token | smiles |
Token ID | 21845 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
ried | +0.263 |
ck | +0.236 |
ases | +0.229 |
beh | +0.227 |
aunts | +0.223 |
Neg logit contributions | |
brand | -0.212 |
sky | -0.194 |
truly | -0.194 |
Not | -0.190 |
� | -0.189 |
Token | and |
Token ID | 290 |
Feature activation | -1.72e-01 |
Pos logit contributions | |
ried | +0.038 |
ck | +0.035 |
ases | +0.034 |
aunts | +0.031 |
ped | +0.030 |
Neg logit contributions | |
brand | -0.052 |
� | -0.048 |
truly | -0.047 |
breathtaking | -0.046 |
Not | -0.046 |
Token | cl |
Token ID | 537 |
Feature activation | +4.23e+00 |
Pos logit contributions | |
brand | +0.006 |
sky | +0.005 |
truly | +0.005 |
Not | +0.005 |
� | +0.005 |
Neg logit contributions | |
ried | -0.005 |
ck | -0.004 |
ases | -0.004 |
aunts | -0.004 |
beh | -0.004 |
Token | aps |
Token ID | 1686 |
Feature activation | +5.19e+00 |
Pos logit contributions | |
ried | +0.118 |
ck | +0.107 |
steer | +0.091 |
ped | +0.090 |
ie | +0.089 |
Neg logit contributions | |
brand | -0.145 |
� | -0.141 |
truly | -0.135 |
Not | -0.135 |
sky | -0.134 |
Token | had |
Token ID | 550 |
Feature activation | +2.64e-01 |
Pos logit contributions | |
brand | +0.003 |
truly | +0.003 |
sky | +0.003 |
� | +0.003 |
Imagine | +0.003 |
Neg logit contributions | |
ried | -0.004 |
ases | -0.003 |
ck | -0.003 |
ucks | -0.003 |
aunts | -0.003 |
Token | fun |
Token ID | 1257 |
Feature activation | -6.41e+00 |
Pos logit contributions | |
ried | +0.010 |
ck | +0.010 |
ases | +0.009 |
share | +0.009 |
his | +0.009 |
Neg logit contributions | |
� | -0.006 |
brand | -0.005 |
Finally | -0.005 |
Imagine | -0.005 |
sky | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | +2.83e+00 |
Pos logit contributions | |
brand | +0.307 |
truly | +0.288 |
� | +0.287 |
breathtaking | +0.281 |
Imagine | +0.281 |
Neg logit contributions | |
ried | -0.096 |
ck | -0.084 |
ases | -0.080 |
aunts | -0.064 |
ies | -0.058 |
Token |
|
Token ID | 198 |
Feature activation | -3.34e+00 |
Pos logit contributions | |
ried | +0.069 |
ases | +0.066 |
ck | +0.063 |
ucks | +0.062 |
aunts | +0.059 |
Neg logit contributions | |
brand | -0.110 |
truly | -0.096 |
breathtaking | -0.095 |
� | -0.092 |
fact | -0.091 |
Token |
|
Token ID | 198 |
Feature activation | -3.39e+00 |
Pos logit contributions | |
brand | +0.084 |
� | +0.076 |
truly | +0.073 |
breathtaking | +0.071 |
Imagine | +0.069 |
Neg logit contributions | |
ried | -0.125 |
ck | -0.122 |
ases | -0.117 |
aunts | -0.111 |
ucks | -0.107 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.52e+01 |
Pos logit contributions | |
brand | +0.095 |
� | +0.084 |
truly | +0.081 |
breathtaking | +0.079 |
sky | +0.076 |
Neg logit contributions | |
ried | -0.122 |
ck | -0.120 |
ases | -0.112 |
aunts | -0.108 |
steer | -0.103 |
Token | Lily |
Token ID | 20037 |
Feature activation | +8.41e+00 |
Pos logit contributions | |
ck | +0.392 |
ried | +0.376 |
ases | +0.338 |
steer | +0.330 |
race | +0.327 |
Neg logit contributions | |
sky | -0.504 |
brand | -0.501 |
� | -0.496 |
Imagine | -0.488 |
truly | -0.484 |
Token | saw |
Token ID | 2497 |
Feature activation | +4.20e+00 |
Pos logit contributions | |
ried | +0.283 |
ases | +0.266 |
ck | +0.265 |
beh | +0.248 |
aunts | +0.242 |
Neg logit contributions | |
brand | -0.241 |
truly | -0.227 |
sky | -0.222 |
� | -0.217 |
Imagine | -0.215 |
Token | them |
Token ID | 606 |
Feature activation | +4.62e-01 |
Pos logit contributions | |
ried | +0.094 |
ck | +0.085 |
ases | +0.081 |
aunts | +0.072 |
race | +0.067 |
Neg logit contributions | |
brand | -0.172 |
� | -0.161 |
truly | -0.160 |
sky | -0.156 |
breathtaking | -0.155 |
Token | and |
Token ID | 290 |
Feature activation | -9.76e-02 |
Pos logit contributions | |
ried | +0.015 |
ck | +0.013 |
ases | +0.013 |
aunts | +0.012 |
beh | +0.012 |
Neg logit contributions | |
brand | -0.014 |
truly | -0.013 |
sky | -0.013 |
� | -0.013 |
Not | -0.013 |
Token | smiled |
Token ID | 13541 |
Feature activation | +2.72e+00 |
Pos logit contributions | |
brand | +0.003 |
truly | +0.003 |
sky | +0.003 |
Imagine | +0.003 |
Not | +0.003 |
Neg logit contributions | |
ried | -0.003 |
ases | -0.003 |
ck | -0.003 |
aunts | -0.003 |
race | -0.003 |
Token | sorry |
Token ID | 7926 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
brand | +0.080 |
truly | +0.078 |
much | +0.065 |
Not | +0.064 |
fantastic | +0.064 |
Neg logit contributions | |
ried | -0.151 |
ases | -0.140 |
race | -0.138 |
talk | -0.135 |
aunts | -0.134 |
Token | again |
Token ID | 757 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
ried | +0.050 |
ck | +0.046 |
ases | +0.044 |
aunts | +0.040 |
race | +0.039 |
Neg logit contributions | |
brand | -0.068 |
truly | -0.063 |
Not | -0.062 |
sky | -0.062 |
� | -0.061 |
Token | . |
Token ID | 13 |
Feature activation | +2.67e+00 |
Pos logit contributions | |
brand | +0.058 |
truly | +0.055 |
� | +0.054 |
sky | +0.054 |
Not | +0.053 |
Neg logit contributions | |
ried | -0.032 |
ck | -0.028 |
ases | -0.027 |
aunts | -0.024 |
beh | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | -3.65e+00 |
Pos logit contributions | |
ried | +0.085 |
ck | +0.078 |
ases | +0.077 |
aunts | +0.072 |
ucks | +0.070 |
Neg logit contributions | |
brand | -0.082 |
truly | -0.075 |
� | -0.073 |
breathtaking | -0.072 |
Imagine | -0.071 |
Token |
|
Token ID | 198 |
Feature activation | -3.71e+00 |
Pos logit contributions | |
brand | +0.072 |
� | +0.062 |
truly | +0.061 |
sky | +0.059 |
Not | +0.059 |
Neg logit contributions | |
ried | -0.160 |
ck | -0.149 |
ases | -0.147 |
aunts | -0.138 |
beh | -0.137 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.51e+01 |
Pos logit contributions | |
brand | +0.112 |
� | +0.102 |
Finally | +0.101 |
truly | +0.101 |
Imagine | +0.100 |
Neg logit contributions | |
ried | -0.126 |
ases | -0.111 |
ck | -0.110 |
beh | -0.101 |
ped | -0.100 |
Token | Lee |
Token ID | 5741 |
Feature activation | +7.31e+00 |
Pos logit contributions | |
ck | +0.403 |
ried | +0.363 |
ases | +0.326 |
steer | +0.318 |
race | +0.302 |
Neg logit contributions | |
brand | -0.533 |
� | -0.514 |
sky | -0.506 |
Imagine | -0.503 |
Not | -0.501 |
Token | smiled |
Token ID | 13541 |
Feature activation | +2.58e+00 |
Pos logit contributions | |
ried | +0.209 |
ck | +0.190 |
ases | +0.189 |
aunts | +0.171 |
beh | +0.166 |
Neg logit contributions | |
brand | -0.253 |
truly | -0.232 |
� | -0.232 |
sky | -0.225 |
Imagine | -0.223 |
Token | . |
Token ID | 13 |
Feature activation | +2.69e+00 |
Pos logit contributions | |
ried | +0.080 |
ck | +0.076 |
ases | +0.073 |
beh | +0.067 |
aunts | +0.067 |
Neg logit contributions | |
brand | -0.085 |
sky | -0.077 |
truly | -0.076 |
Imagine | -0.073 |
� | -0.073 |
Token | She |
Token ID | 1375 |
Feature activation | +1.46e+00 |
Pos logit contributions | |
ried | +0.083 |
ases | +0.076 |
ck | +0.074 |
aunts | +0.071 |
ucks | +0.069 |
Neg logit contributions | |
brand | -0.082 |
truly | -0.077 |
� | -0.076 |
breathtaking | -0.074 |
Imagine | -0.071 |
Token | gave |
Token ID | 2921 |
Feature activation | +7.44e+00 |
Pos logit contributions | |
ried | +0.042 |
ck | +0.039 |
ases | +0.037 |
aunts | +0.034 |
race | +0.033 |
Neg logit contributions | |
brand | -0.050 |
� | -0.046 |
truly | -0.046 |
sky | -0.045 |
Imagine | -0.044 |
Token | was |
Token ID | 373 |
Feature activation | -3.71e+00 |
Pos logit contributions | |
brand | +0.346 |
� | +0.327 |
sky | +0.304 |
truly | +0.302 |
Imagine | +0.301 |
Neg logit contributions | |
ried | -0.443 |
ck | -0.416 |
ases | -0.397 |
aunts | -0.370 |
beh | -0.363 |
Token | wrong |
Token ID | 2642 |
Feature activation | -3.54e+00 |
Pos logit contributions | |
brand | +0.106 |
� | +0.098 |
truly | +0.095 |
sky | +0.092 |
Imagine | +0.092 |
Neg logit contributions | |
ried | -0.127 |
ck | -0.119 |
ases | -0.115 |
aunts | -0.107 |
race | -0.103 |
Token | . |
Token ID | 13 |
Feature activation | +2.32e+00 |
Pos logit contributions | |
brand | +0.150 |
� | +0.140 |
truly | +0.140 |
sky | +0.136 |
Not | +0.135 |
Neg logit contributions | |
ried | -0.073 |
ck | -0.065 |
ases | -0.062 |
aunts | -0.055 |
race | -0.051 |
Token |
|
Token ID | 198 |
Feature activation | -4.00e+00 |
Pos logit contributions | |
ried | +0.072 |
ases | +0.065 |
aunts | +0.064 |
ck | +0.063 |
race | +0.062 |
Neg logit contributions | |
brand | -0.070 |
breathtaking | -0.062 |
truly | -0.061 |
� | -0.059 |
fact | -0.058 |
Token |
|
Token ID | 198 |
Feature activation | -4.08e+00 |
Pos logit contributions | |
brand | +0.099 |
� | +0.086 |
truly | +0.085 |
breathtaking | +0.082 |
sky | +0.081 |
Neg logit contributions | |
ried | -0.150 |
ck | -0.144 |
ases | -0.141 |
aunts | -0.133 |
race | -0.130 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.51e+01 |
Pos logit contributions | |
brand | +0.140 |
� | +0.129 |
truly | +0.127 |
sky | +0.123 |
breathtaking | +0.122 |
Neg logit contributions | |
ried | -0.117 |
ck | -0.109 |
ases | -0.105 |
aunts | -0.097 |
race | -0.094 |
Token | Joyce |
Token ID | 25936 |
Feature activation | +6.01e+00 |
Pos logit contributions | |
ck | +0.451 |
ried | +0.441 |
ases | +0.409 |
steer | +0.402 |
aunts | +0.384 |
Neg logit contributions | |
brand | -0.454 |
truly | -0.441 |
Imagine | -0.432 |
� | -0.429 |
Not | -0.428 |
Token | was |
Token ID | 373 |
Feature activation | -3.63e+00 |
Pos logit contributions | |
ried | +0.219 |
ases | +0.209 |
ck | +0.202 |
aunts | +0.191 |
ucks | +0.190 |
Neg logit contributions | |
brand | -0.152 |
truly | -0.147 |
sky | -0.137 |
Imagine | -0.137 |
� | -0.135 |
Token | so |
Token ID | 523 |
Feature activation | -1.16e+01 |
Pos logit contributions | |
brand | +0.075 |
� | +0.070 |
sky | +0.064 |
Not | +0.062 |
Imagine | +0.062 |
Neg logit contributions | |
ried | -0.150 |
ck | -0.144 |
ases | -0.137 |
aunts | -0.129 |
beh | -0.128 |
Token | sorry |
Token ID | 7926 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
brand | +0.353 |
� | +0.329 |
sky | +0.312 |
truly | +0.312 |
breathtaking | +0.307 |
Neg logit contributions | |
ried | -0.376 |
ck | -0.353 |
ases | -0.339 |
aunts | -0.317 |
beh | -0.308 |
Token | that |
Token ID | 326 |
Feature activation | -3.12e+00 |
Pos logit contributions | |
ried | +0.055 |
ck | +0.051 |
ases | +0.049 |
aunts | +0.045 |
ies | +0.044 |
Neg logit contributions | |
brand | -0.060 |
� | -0.056 |
truly | -0.054 |
breathtaking | -0.053 |
Imagine | -0.053 |
Token | many |
Token ID | 867 |
Feature activation | -3.38e+00 |
Pos logit contributions | |
brand | +0.026 |
truly | +0.024 |
� | +0.023 |
sky | +0.023 |
Not | +0.023 |
Neg logit contributions | |
ried | -0.030 |
ck | -0.028 |
ases | -0.027 |
aunts | -0.025 |
race | -0.024 |
Token | things |
Token ID | 1243 |
Feature activation | -9.42e+00 |
Pos logit contributions | |
brand | +0.113 |
� | +0.110 |
truly | +0.102 |
Imagine | +0.101 |
breathtaking | +0.099 |
Neg logit contributions | |
ried | -0.097 |
ases | -0.094 |
ck | -0.094 |
ies | -0.082 |
aunts | -0.079 |
Token | . |
Token ID | 13 |
Feature activation | +2.73e+00 |
Pos logit contributions | |
brand | +0.349 |
� | +0.322 |
truly | +0.322 |
sky | +0.313 |
Not | +0.311 |
Neg logit contributions | |
ried | -0.245 |
ck | -0.223 |
ases | -0.215 |
aunts | -0.196 |
race | -0.185 |
Token |
|
Token ID | 198 |
Feature activation | -3.44e+00 |
Pos logit contributions | |
ried | +0.075 |
ases | +0.072 |
ck | +0.067 |
aunts | +0.065 |
ucks | +0.064 |
Neg logit contributions | |
brand | -0.092 |
truly | -0.084 |
breathtaking | -0.083 |
� | -0.083 |
Imagine | -0.079 |
Token |
|
Token ID | 198 |
Feature activation | -3.51e+00 |
Pos logit contributions | |
brand | +0.094 |
� | +0.082 |
truly | +0.080 |
breathtaking | +0.079 |
sky | +0.076 |
Neg logit contributions | |
ried | -0.118 |
ck | -0.116 |
ases | -0.113 |
aunts | -0.108 |
ucks | -0.104 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.51e+01 |
Pos logit contributions | |
brand | +0.124 |
truly | +0.115 |
� | +0.115 |
Imagine | +0.113 |
sky | +0.112 |
Neg logit contributions | |
ried | -0.098 |
ck | -0.088 |
ases | -0.086 |
aunts | -0.078 |
ped | -0.075 |
Token | Lily |
Token ID | 20037 |
Feature activation | +8.17e+00 |
Pos logit contributions | |
ck | +0.395 |
ried | +0.389 |
ases | +0.338 |
steer | +0.324 |
aunts | +0.316 |
Neg logit contributions | |
brand | -0.533 |
sky | -0.504 |
Not | -0.501 |
� | -0.498 |
truly | -0.496 |
Token | says |
Token ID | 1139 |
Feature activation | +5.44e+00 |
Pos logit contributions | |
ried | +0.284 |
ck | +0.273 |
ases | +0.252 |
beh | +0.248 |
aunts | +0.244 |
Neg logit contributions | |
brand | -0.218 |
sky | -0.215 |
truly | -0.205 |
Not | -0.202 |
� | -0.199 |
Token | , |
Token ID | 11 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
ried | +0.204 |
ck | +0.192 |
ases | +0.184 |
aunts | +0.177 |
beh | +0.172 |
Neg logit contributions | |
brand | -0.144 |
� | -0.128 |
truly | -0.127 |
Not | -0.126 |
sky | -0.125 |
Token | " |
Token ID | 366 |
Feature activation | -2.85e+00 |
Pos logit contributions | |
brand | +0.034 |
� | +0.030 |
truly | +0.030 |
sky | +0.028 |
breathtaking | +0.028 |
Neg logit contributions | |
ried | -0.067 |
ck | -0.063 |
ases | -0.061 |
aunts | -0.058 |
race | -0.056 |
Token | Do |
Token ID | 5211 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
brand | +0.085 |
� | +0.078 |
truly | +0.078 |
sky | +0.076 |
Imagine | +0.075 |
Neg logit contributions | |
ried | -0.095 |
ck | -0.088 |
ases | -0.086 |
aunts | -0.079 |
ped | -0.076 |
Token | made |
Token ID | 925 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
ried | +0.103 |
ases | +0.099 |
ck | +0.097 |
share | +0.089 |
aunts | +0.085 |
Neg logit contributions | |
brand | -0.101 |
� | -0.091 |
breathtaking | -0.087 |
sky | -0.086 |
truly | -0.086 |
Token | triangles |
Token ID | 44360 |
Feature activation | -3.27e+00 |
Pos logit contributions | |
ck | +0.033 |
ried | +0.033 |
ases | +0.032 |
aunts | +0.028 |
ies | +0.027 |
Neg logit contributions | |
� | -0.038 |
brand | -0.037 |
Imagine | -0.035 |
truly | -0.034 |
breathtaking | -0.033 |
Token | !" |
Token ID | 2474 |
Feature activation | +2.18e+00 |
Pos logit contributions | |
brand | +0.120 |
� | +0.111 |
truly | +0.110 |
Imagine | +0.107 |
sky | +0.107 |
Neg logit contributions | |
ried | -0.086 |
ck | -0.079 |
ases | -0.076 |
aunts | -0.068 |
race | -0.065 |
Token |
|
Token ID | 198 |
Feature activation | -3.38e+00 |
Pos logit contributions | |
ried | +0.077 |
ck | +0.073 |
ases | +0.072 |
aunts | +0.067 |
ucks | +0.064 |
Neg logit contributions | |
brand | -0.059 |
truly | -0.053 |
� | -0.053 |
breathtaking | -0.051 |
Imagine | -0.051 |
Token |
|
Token ID | 198 |
Feature activation | -3.44e+00 |
Pos logit contributions | |
brand | +0.064 |
truly | +0.054 |
� | +0.053 |
sky | +0.053 |
Not | +0.052 |
Neg logit contributions | |
ried | -0.153 |
ck | -0.140 |
ases | -0.138 |
beh | -0.131 |
aunts | -0.130 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.50e+01 |
Pos logit contributions | |
brand | +0.121 |
Finally | +0.113 |
truly | +0.113 |
sky | +0.112 |
Imagine | +0.112 |
Neg logit contributions | |
ried | -0.099 |
ck | -0.084 |
ases | -0.083 |
beh | -0.075 |
ped | -0.075 |
Token | Lily |
Token ID | 20037 |
Feature activation | +8.24e+00 |
Pos logit contributions | |
ck | +0.394 |
ried | +0.384 |
ases | +0.324 |
steer | +0.319 |
beh | +0.314 |
Neg logit contributions | |
brand | -0.519 |
sky | -0.500 |
� | -0.498 |
Not | -0.498 |
Imagine | -0.485 |
Token | smiles |
Token ID | 21845 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
ried | +0.245 |
ck | +0.224 |
ases | +0.212 |
aunts | +0.198 |
beh | +0.195 |
Neg logit contributions | |
brand | -0.276 |
� | -0.254 |
truly | -0.253 |
sky | -0.252 |
Not | -0.246 |
Token | and |
Token ID | 290 |
Feature activation | +2.82e-04 |
Pos logit contributions | |
ried | +0.024 |
ck | +0.022 |
ases | +0.022 |
aunts | +0.020 |
ies | +0.019 |
Neg logit contributions | |
brand | -0.040 |
� | -0.038 |
truly | -0.037 |
breathtaking | -0.036 |
Imagine | -0.035 |
Token | cl |
Token ID | 537 |
Feature activation | +4.10e+00 |
Pos logit contributions | |
ried | +0.000 |
ck | +0.000 |
ases | +0.000 |
aunts | +0.000 |
beh | +0.000 |
Neg logit contributions | |
brand | -0.000 |
truly | -0.000 |
� | -0.000 |
sky | -0.000 |
Not | -0.000 |
Token | aps |
Token ID | 1686 |
Feature activation | +4.86e+00 |
Pos logit contributions | |
ried | +0.114 |
ck | +0.103 |
ped | +0.087 |
ie | +0.086 |
steer | +0.085 |
Neg logit contributions | |
brand | -0.142 |
� | -0.139 |
sky | -0.135 |
Not | -0.134 |
truly | -0.133 |
Token | pet |
Token ID | 4273 |
Feature activation | +9.21e+00 |
Pos logit contributions | |
ried | +0.083 |
ck | +0.074 |
ases | +0.073 |
ped | +0.070 |
aunts | +0.068 |
Neg logit contributions | |
truly | -0.096 |
brand | -0.094 |
� | -0.092 |
Not | -0.088 |
definitely | -0.086 |
Token | him |
Token ID | 683 |
Feature activation | +3.32e+00 |
Pos logit contributions | |
ried | +0.304 |
ck | +0.274 |
pered | +0.244 |
ases | +0.243 |
beh | +0.238 |
Neg logit contributions | |
brand | -0.279 |
sky | -0.276 |
Not | -0.271 |
� | -0.254 |
lot | -0.252 |
Token | ?" |
Token ID | 1701 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
ried | +0.104 |
ck | +0.097 |
ases | +0.093 |
aunts | +0.087 |
beh | +0.086 |
Neg logit contributions | |
brand | -0.108 |
truly | -0.098 |
� | -0.098 |
sky | -0.096 |
Not | -0.094 |
Token |
|
Token ID | 198 |
Feature activation | -3.45e+00 |
Pos logit contributions | |
ried | +0.047 |
ck | +0.044 |
ases | +0.043 |
beh | +0.040 |
aunts | +0.040 |
Neg logit contributions | |
brand | -0.022 |
� | -0.020 |
Not | -0.020 |
sky | -0.020 |
truly | -0.019 |
Token |
|
Token ID | 198 |
Feature activation | -3.51e+00 |
Pos logit contributions | |
brand | +0.049 |
Not | +0.041 |
� | +0.041 |
sky | +0.041 |
Finally | +0.039 |
Neg logit contributions | |
ried | -0.177 |
ck | -0.159 |
ases | -0.158 |
beh | -0.155 |
rying | -0.151 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.50e+01 |
Pos logit contributions | |
brand | +0.110 |
Finally | +0.107 |
Imagine | +0.105 |
Not | +0.103 |
truly | +0.101 |
Neg logit contributions | |
ried | -0.118 |
ases | -0.099 |
ck | -0.098 |
ped | -0.096 |
rying | -0.091 |
Token | Rose |
Token ID | 8049 |
Feature activation | +7.23e+00 |
Pos logit contributions | |
ck | +0.449 |
ried | +0.424 |
ases | +0.380 |
steer | +0.371 |
beh | +0.365 |
Neg logit contributions | |
brand | -0.484 |
� | -0.450 |
sky | -0.448 |
Imagine | -0.446 |
Not | -0.445 |
Token | nodded |
Token ID | 14464 |
Feature activation | +4.07e+00 |
Pos logit contributions | |
ried | +0.212 |
ck | +0.194 |
ases | +0.193 |
aunts | +0.176 |
beh | +0.170 |
Neg logit contributions | |
brand | -0.244 |
truly | -0.224 |
� | -0.224 |
sky | -0.218 |
Imagine | -0.215 |
Token | and |
Token ID | 290 |
Feature activation | -6.74e-02 |
Pos logit contributions | |
ried | +0.152 |
ases | +0.145 |
beh | +0.143 |
ck | +0.142 |
race | +0.135 |
Neg logit contributions | |
brand | -0.108 |
sky | -0.098 |
truly | -0.093 |
Imagine | -0.090 |
so | -0.086 |
Token | said |
Token ID | 531 |
Feature activation | +5.78e+00 |
Pos logit contributions | |
brand | +0.002 |
truly | +0.002 |
sky | +0.002 |
Not | +0.002 |
Imagine | +0.002 |
Neg logit contributions | |
ried | -0.002 |
ases | -0.002 |
aunts | -0.002 |
ck | -0.002 |
beh | -0.002 |
Token | , |
Token ID | 11 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
ried | +0.234 |
ck | +0.224 |
ases | +0.217 |
aunts | +0.213 |
beh | +0.206 |
Neg logit contributions | |
brand | -0.135 |
� | -0.114 |
Not | -0.113 |
truly | -0.112 |
sky | -0.106 |
Token | new |
Token ID | 649 |
Feature activation | -2.66e+00 |
Pos logit contributions | |
brand | +0.007 |
� | +0.007 |
truly | +0.007 |
Imagine | +0.007 |
Not | +0.007 |
Neg logit contributions | |
ried | -0.006 |
ck | -0.005 |
ases | -0.005 |
aunts | -0.005 |
race | -0.005 |
Token | toy |
Token ID | 13373 |
Feature activation | -3.92e+00 |
Pos logit contributions | |
brand | +0.089 |
truly | +0.080 |
� | +0.080 |
sky | +0.079 |
breathtaking | +0.077 |
Neg logit contributions | |
ried | -0.081 |
ck | -0.074 |
ases | -0.072 |
aunts | -0.066 |
beh | -0.064 |
Token | . |
Token ID | 13 |
Feature activation | +2.66e+00 |
Pos logit contributions | |
brand | +0.154 |
sky | +0.143 |
� | +0.143 |
truly | +0.142 |
Not | +0.138 |
Neg logit contributions | |
ried | -0.093 |
ck | -0.086 |
ases | -0.078 |
aunts | -0.073 |
beh | -0.071 |
Token |
|
Token ID | 198 |
Feature activation | -3.56e+00 |
Pos logit contributions | |
ried | +0.080 |
ases | +0.076 |
ck | +0.073 |
aunts | +0.072 |
ucks | +0.071 |
Neg logit contributions | |
brand | -0.082 |
truly | -0.075 |
� | -0.075 |
breathtaking | -0.072 |
fact | -0.070 |
Token |
|
Token ID | 198 |
Feature activation | -3.56e+00 |
Pos logit contributions | |
brand | +0.075 |
� | +0.065 |
truly | +0.065 |
sky | +0.062 |
Not | +0.061 |
Neg logit contributions | |
ried | -0.149 |
ck | -0.141 |
ases | -0.138 |
aunts | -0.130 |
race | -0.127 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.50e+01 |
Pos logit contributions | |
brand | +0.107 |
� | +0.097 |
truly | +0.097 |
Not | +0.096 |
Imagine | +0.096 |
Neg logit contributions | |
ried | -0.118 |
ck | -0.107 |
ases | -0.104 |
ped | -0.095 |
beh | -0.094 |
Token | Jen |
Token ID | 13364 |
Feature activation | +8.74e+00 |
Pos logit contributions | |
ck | +0.378 |
ried | +0.361 |
steer | +0.315 |
beh | +0.309 |
race | +0.300 |
Neg logit contributions | |
sky | -0.515 |
brand | -0.510 |
� | -0.490 |
Not | -0.487 |
truly | -0.481 |
Token | hugs |
Token ID | 40657 |
Feature activation | +7.64e+00 |
Pos logit contributions | |
ried | +0.283 |
ck | +0.259 |
ases | +0.247 |
aunts | +0.235 |
beh | +0.233 |
Neg logit contributions | |
brand | -0.268 |
sky | -0.246 |
truly | -0.246 |
� | -0.245 |
Not | -0.240 |
Token | them |
Token ID | 606 |
Feature activation | +3.76e-01 |
Pos logit contributions | |
ried | +0.211 |
ck | +0.188 |
beh | +0.180 |
ases | +0.164 |
maneu | +0.159 |
Neg logit contributions | |
sky | -0.267 |
brand | -0.247 |
Not | -0.244 |
� | -0.242 |
lot | -0.237 |
Token | and |
Token ID | 290 |
Feature activation | -4.48e-03 |
Pos logit contributions | |
ried | +0.013 |
ck | +0.013 |
ases | +0.012 |
beh | +0.011 |
aunts | +0.011 |
Neg logit contributions | |
brand | -0.011 |
sky | -0.011 |
much | -0.010 |
so | -0.010 |
very | -0.010 |
Token | says |
Token ID | 1139 |
Feature activation | +5.56e+00 |
Pos logit contributions | |
brand | +0.000 |
� | +0.000 |
truly | +0.000 |
sky | +0.000 |
Not | +0.000 |
Neg logit contributions | |
ried | -0.000 |
ck | -0.000 |
ases | -0.000 |
aunts | -0.000 |
race | -0.000 |
Token | . |
Token ID | 13 |
Feature activation | +2.14e+00 |
Pos logit contributions | |
ried | +0.149 |
ck | +0.133 |
ases | +0.125 |
aunts | +0.120 |
beh | +0.116 |
Neg logit contributions | |
brand | -0.151 |
� | -0.141 |
sky | -0.139 |
truly | -0.137 |
Not | -0.137 |
Token | Dog |
Token ID | 8532 |
Feature activation | +1.27e+01 |
Pos logit contributions | |
ried | +0.087 |
ck | +0.087 |
ases | +0.078 |
ies | +0.078 |
ie | +0.077 |
Neg logit contributions | |
Not | -0.048 |
brand | -0.047 |
� | -0.044 |
sky | -0.043 |
Cookies | -0.042 |
Token | ." |
Token ID | 526 |
Feature activation | +1.78e+00 |
Pos logit contributions | |
ried | +0.279 |
ck | +0.271 |
ases | +0.268 |
race | +0.254 |
aunts | +0.236 |
Neg logit contributions | |
brand | -0.462 |
Not | -0.449 |
� | -0.436 |
truly | -0.436 |
may | -0.433 |
Token |
|
Token ID | 198 |
Feature activation | -3.93e+00 |
Pos logit contributions | |
ried | +0.078 |
ck | +0.075 |
ases | +0.071 |
beh | +0.067 |
aunts | +0.067 |
Neg logit contributions | |
brand | -0.036 |
� | -0.032 |
sky | -0.032 |
Not | -0.032 |
truly | -0.030 |
Token |
|
Token ID | 198 |
Feature activation | -3.97e+00 |
Pos logit contributions | |
brand | +0.046 |
Finally | +0.040 |
Not | +0.039 |
sky | +0.037 |
� | +0.036 |
Neg logit contributions | |
ried | -0.206 |
ck | -0.186 |
ases | -0.183 |
beh | -0.181 |
rying | -0.176 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.50e+01 |
Pos logit contributions | |
Finally | +0.109 |
Eventually | +0.101 |
Imagine | +0.100 |
brand | +0.099 |
Very | +0.097 |
Neg logit contributions | |
ried | -0.149 |
ck | -0.126 |
ases | -0.126 |
ped | -0.125 |
ie | -0.122 |
Token | Lee |
Token ID | 5741 |
Feature activation | +7.25e+00 |
Pos logit contributions | |
ck | +0.468 |
ried | +0.415 |
steer | +0.373 |
ases | +0.368 |
aunts | +0.352 |
Neg logit contributions | |
Not | -0.457 |
� | -0.453 |
brand | -0.438 |
sky | -0.430 |
Very | -0.426 |
Token | says |
Token ID | 1139 |
Feature activation | +5.31e+00 |
Pos logit contributions | |
ried | +0.241 |
beh | +0.209 |
ases | +0.208 |
ck | +0.208 |
aunts | +0.199 |
Neg logit contributions | |
brand | -0.215 |
sky | -0.205 |
Not | -0.200 |
� | -0.199 |
truly | -0.198 |
Token | , |
Token ID | 11 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
ried | +0.243 |
ck | +0.230 |
ases | +0.225 |
aunts | +0.222 |
beh | +0.218 |
Neg logit contributions | |
brand | -0.099 |
� | -0.090 |
Not | -0.083 |
much | -0.080 |
Cookies | -0.078 |
Token | " |
Token ID | 366 |
Feature activation | -2.85e+00 |
Pos logit contributions | |
brand | +0.038 |
truly | +0.034 |
� | +0.033 |
breathtaking | +0.032 |
Imagine | +0.031 |
Neg logit contributions | |
ried | -0.067 |
ck | -0.062 |
ases | -0.061 |
aunts | -0.058 |
race | -0.056 |
Token | Very |
Token ID | 16371 |
Feature activation | -9.11e+00 |
Pos logit contributions | |
brand | +0.076 |
� | +0.069 |
truly | +0.068 |
sky | +0.066 |
Imagine | +0.065 |
Neg logit contributions | |
ried | -0.104 |
ck | -0.097 |
ases | -0.095 |
aunts | -0.088 |
race | -0.086 |
Token | get |
Token ID | 651 |
Feature activation | -1.65e+00 |
Pos logit contributions | |
brand | +0.011 |
� | +0.010 |
truly | +0.010 |
sky | +0.010 |
breathtaking | +0.010 |
Neg logit contributions | |
ried | -0.007 |
ck | -0.007 |
ases | -0.007 |
aunts | -0.006 |
race | -0.006 |
Token | hurt |
Token ID | 5938 |
Feature activation | +3.34e+00 |
Pos logit contributions | |
brand | +0.054 |
truly | +0.049 |
sky | +0.047 |
much | +0.046 |
most | +0.044 |
Neg logit contributions | |
ried | -0.049 |
ases | -0.047 |
aunts | -0.044 |
ck | -0.044 |
ucks | -0.042 |
Token | ." |
Token ID | 526 |
Feature activation | +2.11e+00 |
Pos logit contributions | |
ried | +0.096 |
ck | +0.088 |
ases | +0.080 |
aunts | +0.079 |
maneu | +0.077 |
Neg logit contributions | |
brand | -0.109 |
sky | -0.107 |
truly | -0.103 |
Imagine | -0.103 |
Not | -0.101 |
Token |
|
Token ID | 198 |
Feature activation | -3.62e+00 |
Pos logit contributions | |
ried | +0.106 |
ck | +0.098 |
ases | +0.095 |
beh | +0.094 |
rying | +0.092 |
Neg logit contributions | |
Not | -0.033 |
� | -0.028 |
sky | -0.027 |
Cookies | -0.026 |
Once | -0.026 |
Token |
|
Token ID | 198 |
Feature activation | -3.79e+00 |
Pos logit contributions | |
brand | +0.036 |
Not | +0.034 |
Finally | +0.034 |
sky | +0.031 |
Imagine | +0.030 |
Neg logit contributions | |
ried | -0.196 |
ases | -0.174 |
beh | -0.173 |
ck | -0.171 |
rying | -0.167 |
Token | Jes |
Token ID | 22290 |
Feature activation | +1.50e+01 |
Pos logit contributions | |
Finally | +0.102 |
Imagine | +0.095 |
Not | +0.094 |
Eventually | +0.093 |
Very | +0.090 |
Neg logit contributions | |
ried | -0.144 |
ases | -0.124 |
ped | -0.121 |
ck | -0.119 |
leave | -0.117 |
Token | se |
Token ID | 325 |
Feature activation | +1.16e+01 |
Pos logit contributions | |
ck | +0.364 |
ried | +0.327 |
ases | +0.320 |
ie | +0.311 |
aunts | +0.307 |
Neg logit contributions | |
brand | -0.515 |
Not | -0.500 |
� | -0.495 |
sky | -0.495 |
may | -0.493 |
Token | was |
Token ID | 373 |
Feature activation | -3.57e+00 |
Pos logit contributions | |
ried | +0.412 |
ases | +0.386 |
beh | +0.383 |
aunts | +0.381 |
ck | +0.381 |
Neg logit contributions | |
fact | -0.242 |
sky | -0.233 |
Imagine | -0.231 |
truly | -0.221 |
Too | -0.216 |
Token | sad |
Token ID | 6507 |
Feature activation | -4.58e+00 |
Pos logit contributions | |
brand | +0.090 |
truly | +0.080 |
� | +0.077 |
sky | +0.076 |
breathtaking | +0.075 |
Neg logit contributions | |
ried | -0.137 |
ck | -0.127 |
ases | -0.126 |
aunts | -0.118 |
race | -0.115 |
Token | . |
Token ID | 13 |
Feature activation | +2.05e+00 |
Pos logit contributions | |
brand | +0.170 |
� | +0.159 |
truly | +0.158 |
breathtaking | +0.154 |
Imagine | +0.153 |
Neg logit contributions | |
ried | -0.119 |
ck | -0.109 |
ases | -0.105 |
aunts | -0.094 |
ies | -0.090 |
Token | He |
Token ID | 679 |
Feature activation | +2.72e-01 |
Pos logit contributions | |
ried | +0.068 |
ck | +0.063 |
ases | +0.061 |
aunts | +0.057 |
beh | +0.055 |
Neg logit contributions | |
brand | -0.062 |
� | -0.056 |
truly | -0.056 |
sky | -0.054 |
Not | -0.054 |
Token | hurting |
Token ID | 21530 |
Feature activation | +3.90e+00 |
Pos logit contributions | |
brand | +0.050 |
truly | +0.046 |
much | +0.044 |
so | +0.043 |
fun | +0.042 |
Neg logit contributions | |
ried | -0.075 |
ucks | -0.068 |
aunts | -0.068 |
ases | -0.067 |
talk | -0.067 |
Token | anyone |
Token ID | 2687 |
Feature activation | -1.78e+00 |
Pos logit contributions | |
ried | +0.112 |
ck | +0.097 |
ases | +0.089 |
maneu | +0.087 |
race | +0.086 |
Neg logit contributions | |
brand | -0.133 |
sky | -0.128 |
Not | -0.125 |
truly | -0.122 |
lot | -0.119 |
Token | ." |
Token ID | 526 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
brand | +0.074 |
truly | +0.069 |
� | +0.068 |
sky | +0.068 |
Not | +0.067 |
Neg logit contributions | |
ried | -0.038 |
ck | -0.033 |
ases | -0.032 |
aunts | -0.028 |
race | -0.027 |
Token |
|
Token ID | 198 |
Feature activation | -3.66e+00 |
Pos logit contributions | |
ried | +0.077 |
ck | +0.072 |
ases | +0.071 |
aunts | +0.067 |
race | +0.065 |
Neg logit contributions | |
brand | -0.049 |
truly | -0.043 |
� | -0.043 |
breathtaking | -0.041 |
Imagine | -0.041 |
Token |
|
Token ID | 198 |
Feature activation | -3.74e+00 |
Pos logit contributions | |
brand | +0.066 |
� | +0.056 |
truly | +0.054 |
sky | +0.053 |
Not | +0.052 |
Neg logit contributions | |
ried | -0.171 |
ck | -0.156 |
ases | -0.154 |
beh | -0.147 |
aunts | -0.145 |
Token | Miss |
Token ID | 17140 |
Feature activation | +1.50e+01 |
Pos logit contributions | |
brand | +0.112 |
Finally | +0.106 |
Imagine | +0.103 |
truly | +0.102 |
Not | +0.099 |
Neg logit contributions | |
ried | -0.131 |
ases | -0.113 |
ck | -0.109 |
ped | -0.106 |
ie | -0.103 |
Token | Lee |
Token ID | 5741 |
Feature activation | +7.21e+00 |
Pos logit contributions | |
ck | +0.412 |
ried | +0.366 |
steer | +0.336 |
ases | +0.310 |
beh | +0.307 |
Neg logit contributions | |
brand | -0.506 |
� | -0.499 |
sky | -0.492 |
Not | -0.488 |
Imagine | -0.476 |
Token | says |
Token ID | 1139 |
Feature activation | +5.43e+00 |
Pos logit contributions | |
ried | +0.246 |
ck | +0.225 |
ases | +0.217 |
aunts | +0.205 |
beh | +0.204 |
Neg logit contributions | |
brand | -0.210 |
� | -0.190 |
truly | -0.189 |
sky | -0.189 |
Not | -0.187 |
Token | , |
Token ID | 11 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
ried | +0.203 |
ck | +0.190 |
ases | +0.185 |
aunts | +0.175 |
beh | +0.170 |
Neg logit contributions | |
brand | -0.144 |
� | -0.130 |
truly | -0.124 |
Not | -0.124 |
sky | -0.122 |
Token | " |
Token ID | 366 |
Feature activation | -2.94e+00 |
Pos logit contributions | |
brand | +0.039 |
truly | +0.035 |
� | +0.033 |
breathtaking | +0.033 |
sky | +0.032 |
Neg logit contributions | |
ried | -0.067 |
ck | -0.062 |
ases | -0.061 |
aunts | -0.058 |
ped | -0.056 |
Token | I |
Token ID | 40 |
Feature activation | +3.71e+00 |
Pos logit contributions | |
brand | +0.082 |
� | +0.071 |
truly | +0.071 |
breathtaking | +0.068 |
sky | +0.067 |
Neg logit contributions | |
ried | -0.106 |
ck | -0.099 |
ases | -0.097 |
aunts | -0.093 |
beh | -0.090 |
Token | mom |
Token ID | 1995 |
Feature activation | +7.46e+00 |
Pos logit contributions | |
ried | +0.071 |
ck | +0.066 |
ases | +0.063 |
beh | +0.056 |
aunts | +0.055 |
Neg logit contributions | |
brand | -0.103 |
truly | -0.094 |
� | -0.093 |
sky | -0.092 |
breathtaking | -0.089 |
Token | my |
Token ID | 1820 |
Feature activation | +4.16e+00 |
Pos logit contributions | |
ases | +0.249 |
ried | +0.237 |
ck | +0.232 |
pered | +0.215 |
beh | +0.209 |
Neg logit contributions | |
brand | -0.223 |
truly | -0.214 |
Imagine | -0.207 |
sky | -0.201 |
definitely | -0.200 |
Token | gave |
Token ID | 2921 |
Feature activation | +7.20e+00 |
Pos logit contributions | |
ases | +0.145 |
ried | +0.135 |
ck | +0.132 |
beh | +0.125 |
ucks | +0.124 |
Neg logit contributions | |
brand | -0.119 |
truly | -0.114 |
Imagine | -0.108 |
definitely | -0.107 |
sky | -0.106 |
Token | him |
Token ID | 683 |
Feature activation | +2.84e+00 |
Pos logit contributions | |
ried | +0.177 |
ck | +0.160 |
ases | +0.159 |
ided | +0.154 |
aunts | +0.150 |
Neg logit contributions | |
brand | -0.256 |
so | -0.242 |
sky | -0.239 |
truly | -0.238 |
Not | -0.237 |
Token | a |
Token ID | 257 |
Feature activation | -4.38e+00 |
Pos logit contributions | |
ried | +0.121 |
ck | +0.116 |
ases | +0.112 |
aunts | +0.110 |
race | +0.103 |
Neg logit contributions | |
brand | -0.064 |
truly | -0.054 |
fantastic | -0.048 |
Not | -0.048 |
much | -0.048 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.74e+01 |
Pos logit contributions | |
� | +0.124 |
brand | +0.122 |
Imagine | +0.120 |
Not | +0.120 |
sky | +0.118 |
Neg logit contributions | |
ried | -0.128 |
ck | -0.119 |
ases | -0.111 |
steer | -0.108 |
leave | -0.107 |
Token | new |
Token ID | 649 |
Feature activation | -3.48e+00 |
Pos logit contributions | |
brand | +0.429 |
� | +0.419 |
Not | +0.401 |
truly | +0.400 |
Imagine | +0.399 |
Neg logit contributions | |
ried | -0.622 |
ck | -0.585 |
ases | -0.568 |
aunts | -0.538 |
ucks | -0.531 |
Token | pair |
Token ID | 5166 |
Feature activation | +2.58e-01 |
Pos logit contributions | |
brand | +0.112 |
� | +0.103 |
truly | +0.102 |
Not | +0.099 |
sky | +0.099 |
Neg logit contributions | |
ried | -0.106 |
ck | -0.097 |
ases | -0.095 |
aunts | -0.087 |
race | -0.084 |
Token | of |
Token ID | 286 |
Feature activation | +5.07e+00 |
Pos logit contributions | |
ried | +0.008 |
ck | +0.008 |
ases | +0.008 |
aunts | +0.007 |
beh | +0.007 |
Neg logit contributions | |
brand | -0.008 |
truly | -0.007 |
� | -0.007 |
sky | -0.007 |
Not | -0.007 |
Token | shoes |
Token ID | 10012 |
Feature activation | -6.40e-01 |
Pos logit contributions | |
ried | +0.179 |
ck | +0.158 |
ases | +0.155 |
aunts | +0.151 |
ucks | +0.142 |
Neg logit contributions | |
brand | -0.160 |
truly | -0.140 |
sky | -0.136 |
breathtaking | -0.131 |
fantastic | -0.130 |
Token | to |
Token ID | 284 |
Feature activation | +5.53e+00 |
Pos logit contributions | |
brand | +0.023 |
truly | +0.021 |
� | +0.021 |
sky | +0.021 |
breathtaking | +0.020 |
Neg logit contributions | |
ried | -0.017 |
ck | -0.016 |
ases | -0.016 |
aunts | -0.014 |
beh | -0.014 |
Token | her |
Token ID | 607 |
Feature activation | +3.29e+00 |
Pos logit contributions | |
ried | +0.096 |
ases | +0.080 |
ck | +0.075 |
aunts | +0.075 |
ucks | +0.070 |
Neg logit contributions | |
brand | -0.185 |
� | -0.177 |
truly | -0.175 |
Not | -0.168 |
sky | -0.167 |
Token | - |
Token ID | 12 |
Feature activation | +1.86e-01 |
Pos logit contributions | |
ried | +0.091 |
ck | +0.082 |
ases | +0.081 |
aunts | +0.075 |
ucks | +0.073 |
Neg logit contributions | |
brand | -0.125 |
truly | -0.113 |
� | -0.111 |
sky | -0.110 |
Not | -0.107 |
Token | there |
Token ID | 612 |
Feature activation | -3.80e+00 |
Pos logit contributions | |
ried | +0.006 |
ases | +0.006 |
ck | +0.006 |
aunts | +0.005 |
pered | +0.005 |
Neg logit contributions | |
brand | -0.006 |
truly | -0.005 |
sky | -0.005 |
Not | -0.005 |
breathtaking | -0.005 |
Token | was |
Token ID | 373 |
Feature activation | -4.30e+00 |
Pos logit contributions | |
brand | +0.163 |
� | +0.153 |
truly | +0.153 |
sky | +0.150 |
Imagine | +0.148 |
Neg logit contributions | |
ried | -0.077 |
ck | -0.068 |
ases | -0.065 |
aunts | -0.057 |
beh | -0.053 |
Token | a |
Token ID | 257 |
Feature activation | -4.49e+00 |
Pos logit contributions | |
brand | +0.128 |
truly | +0.115 |
� | +0.112 |
sky | +0.110 |
breathtaking | +0.109 |
Neg logit contributions | |
ried | -0.149 |
ck | -0.139 |
ases | -0.137 |
aunts | -0.128 |
ucks | -0.123 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.74e+01 |
Pos logit contributions | |
� | +0.116 |
brand | +0.113 |
Not | +0.113 |
truly | +0.108 |
Imagine | +0.107 |
Neg logit contributions | |
ried | -0.134 |
ck | -0.130 |
share | -0.118 |
leave | -0.118 |
ies | -0.117 |
Token | new |
Token ID | 649 |
Feature activation | -3.66e+00 |
Pos logit contributions | |
brand | +0.413 |
� | +0.375 |
truly | +0.367 |
sky | +0.353 |
Imagine | +0.353 |
Neg logit contributions | |
ried | -0.673 |
ck | -0.632 |
ases | -0.616 |
aunts | -0.581 |
beh | -0.562 |
Token | box |
Token ID | 3091 |
Feature activation | -7.59e+00 |
Pos logit contributions | |
brand | +0.108 |
� | +0.100 |
truly | +0.098 |
Not | +0.096 |
Imagine | +0.095 |
Neg logit contributions | |
ried | -0.118 |
ck | -0.110 |
ases | -0.107 |
aunts | -0.099 |
race | -0.097 |
Token | of |
Token ID | 286 |
Feature activation | +4.89e+00 |
Pos logit contributions | |
brand | +0.269 |
truly | +0.249 |
sky | +0.239 |
� | +0.236 |
case | +0.227 |
Neg logit contributions | |
ried | -0.223 |
ases | -0.218 |
ck | -0.210 |
aunts | -0.202 |
ucks | -0.201 |
Token | cr |
Token ID | 1067 |
Feature activation | -3.83e+00 |
Pos logit contributions | |
ried | +0.176 |
ck | +0.161 |
ases | +0.157 |
aunts | +0.148 |
beh | +0.142 |
Neg logit contributions | |
brand | -0.140 |
truly | -0.124 |
sky | -0.121 |
� | -0.121 |
breathtaking | -0.118 |
Token | ay |
Token ID | 323 |
Feature activation | -4.73e-02 |
Pos logit contributions | |
brand | +0.104 |
truly | +0.093 |
� | +0.093 |
Imagine | +0.089 |
breathtaking | +0.089 |
Neg logit contributions | |
ried | -0.136 |
ck | -0.127 |
ases | -0.127 |
aunts | -0.117 |
race | -0.114 |
Token | wall |
Token ID | 3355 |
Feature activation | -7.93e+00 |
Pos logit contributions | |
brand | +0.017 |
� | +0.015 |
truly | +0.015 |
Imagine | +0.014 |
breathtaking | +0.014 |
Neg logit contributions | |
ried | -0.026 |
ck | -0.025 |
ases | -0.024 |
aunts | -0.023 |
race | -0.022 |
Token | , |
Token ID | 11 |
Feature activation | -2.75e+00 |
Pos logit contributions | |
brand | +0.256 |
� | +0.234 |
truly | +0.231 |
breathtaking | +0.224 |
Not | +0.223 |
Neg logit contributions | |
ried | -0.246 |
ck | -0.223 |
ases | -0.219 |
aunts | -0.200 |
race | -0.195 |
Token | there |
Token ID | 612 |
Feature activation | -3.67e+00 |
Pos logit contributions | |
brand | +0.088 |
truly | +0.081 |
� | +0.080 |
sky | +0.078 |
Not | +0.077 |
Neg logit contributions | |
ried | -0.086 |
ck | -0.080 |
ases | -0.078 |
aunts | -0.072 |
beh | -0.069 |
Token | was |
Token ID | 373 |
Feature activation | -4.01e+00 |
Pos logit contributions | |
brand | +0.090 |
� | +0.078 |
truly | +0.076 |
breathtaking | +0.074 |
Not | +0.073 |
Neg logit contributions | |
ried | -0.142 |
ck | -0.133 |
ases | -0.130 |
aunts | -0.123 |
ies | -0.119 |
Token | a |
Token ID | 257 |
Feature activation | -4.38e+00 |
Pos logit contributions | |
brand | +0.133 |
� | +0.131 |
Imagine | +0.125 |
Not | +0.122 |
truly | +0.121 |
Neg logit contributions | |
ried | -0.106 |
ck | -0.092 |
ases | -0.089 |
her | -0.083 |
his | -0.083 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.74e+01 |
Pos logit contributions | |
� | +0.118 |
Not | +0.116 |
brand | +0.114 |
truly | +0.109 |
Imagine | +0.107 |
Neg logit contributions | |
ried | -0.122 |
ck | -0.119 |
share | -0.111 |
ies | -0.111 |
steer | -0.110 |
Token | new |
Token ID | 649 |
Feature activation | -3.46e+00 |
Pos logit contributions | |
brand | +0.398 |
truly | +0.345 |
� | +0.344 |
sky | +0.328 |
breathtaking | +0.324 |
Neg logit contributions | |
ried | -0.703 |
ck | -0.661 |
ases | -0.649 |
aunts | -0.610 |
race | -0.594 |
Token | clock |
Token ID | 8801 |
Feature activation | -3.56e+00 |
Pos logit contributions | |
brand | +0.104 |
� | +0.096 |
truly | +0.094 |
Not | +0.092 |
Imagine | +0.091 |
Neg logit contributions | |
ried | -0.112 |
ck | -0.103 |
ases | -0.102 |
aunts | -0.094 |
ies | -0.092 |
Token | . |
Token ID | 13 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
brand | +0.127 |
sky | +0.121 |
truly | +0.118 |
� | +0.115 |
breathtaking | +0.111 |
Neg logit contributions | |
ried | -0.102 |
ases | -0.092 |
ck | -0.089 |
aunts | -0.085 |
ucks | -0.085 |
Token | It |
Token ID | 632 |
Feature activation | -1.04e+01 |
Pos logit contributions | |
ried | +0.031 |
ases | +0.027 |
pered | +0.023 |
Kim | +0.023 |
race | +0.023 |
Neg logit contributions | |
brand | -0.062 |
truly | -0.057 |
Imagine | -0.056 |
breathtaking | -0.055 |
fact | -0.054 |
Token | was |
Token ID | 373 |
Feature activation | -4.19e+00 |
Pos logit contributions | |
brand | +0.263 |
� | +0.238 |
truly | +0.228 |
Not | +0.222 |
sky | +0.220 |
Neg logit contributions | |
ried | -0.390 |
ck | -0.368 |
ases | -0.353 |
aunts | -0.332 |
ped | -0.322 |
Token | eyes |
Token ID | 2951 |
Feature activation | -5.95e+00 |
Pos logit contributions | |
ried | +0.112 |
ck | +0.104 |
ases | +0.101 |
aunts | +0.095 |
race | +0.094 |
Neg logit contributions | |
brand | -0.093 |
truly | -0.080 |
� | -0.079 |
sky | -0.078 |
breathtaking | -0.076 |
Token | . |
Token ID | 13 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
brand | +0.227 |
� | +0.211 |
truly | +0.209 |
Not | +0.204 |
Imagine | +0.204 |
Neg logit contributions | |
ried | -0.146 |
ck | -0.132 |
ases | -0.124 |
aunts | -0.112 |
race | -0.107 |
Token | Inside |
Token ID | 14384 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
ried | +0.044 |
ases | +0.038 |
race | +0.036 |
pered | +0.036 |
ck | +0.036 |
Neg logit contributions | |
brand | -0.041 |
truly | -0.037 |
breathtaking | -0.036 |
fact | -0.035 |
Imagine | -0.035 |
Token | was |
Token ID | 373 |
Feature activation | -4.47e+00 |
Pos logit contributions | |
brand | +0.036 |
� | +0.032 |
truly | +0.032 |
Not | +0.031 |
Imagine | +0.031 |
Neg logit contributions | |
ried | -0.037 |
ck | -0.034 |
ases | -0.033 |
aunts | -0.031 |
ie | -0.030 |
Token | a |
Token ID | 257 |
Feature activation | -4.65e+00 |
Pos logit contributions | |
brand | +0.134 |
� | +0.125 |
truly | +0.121 |
Imagine | +0.119 |
Not | +0.118 |
Neg logit contributions | |
ried | -0.144 |
ck | -0.134 |
ases | -0.129 |
aunts | -0.119 |
ped | -0.116 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.74e+01 |
Pos logit contributions | |
� | +0.113 |
Not | +0.113 |
Finally | +0.105 |
Eventually | +0.105 |
Too | +0.105 |
Neg logit contributions | |
ried | -0.138 |
ck | -0.137 |
ies | -0.130 |
share | -0.129 |
leave | -0.129 |
Token | new |
Token ID | 649 |
Feature activation | -3.80e+00 |
Pos logit contributions | |
brand | +0.411 |
� | +0.399 |
Imagine | +0.380 |
Not | +0.379 |
truly | +0.377 |
Neg logit contributions | |
ried | -0.643 |
ck | -0.609 |
ases | -0.586 |
aunts | -0.556 |
ies | -0.547 |
Token | dress |
Token ID | 6576 |
Feature activation | -2.05e+00 |
Pos logit contributions | |
brand | +0.114 |
� | +0.107 |
truly | +0.104 |
Not | +0.103 |
Imagine | +0.102 |
Neg logit contributions | |
ried | -0.119 |
ck | -0.111 |
ases | -0.108 |
aunts | -0.100 |
ies | -0.098 |
Token | ! |
Token ID | 0 |
Feature activation | +5.70e-01 |
Pos logit contributions | |
brand | +0.067 |
truly | +0.060 |
� | +0.060 |
sky | +0.059 |
breathtaking | +0.057 |
Neg logit contributions | |
ried | -0.066 |
ck | -0.061 |
ases | -0.061 |
aunts | -0.056 |
beh | -0.054 |
Token | |
Token ID | 220 |
Feature activation | -1.28e+01 |
Pos logit contributions | |
ried | +0.018 |
ases | +0.016 |
ck | +0.015 |
race | +0.015 |
aunts | +0.015 |
Neg logit contributions | |
brand | -0.017 |
truly | -0.015 |
breathtaking | -0.015 |
Imagine | -0.015 |
fact | -0.014 |
Token |
|
Token ID | 198 |
Feature activation | -5.05e+00 |
Pos logit contributions | |
brand | +0.263 |
� | +0.230 |
truly | +0.227 |
sky | +0.215 |
Not | +0.214 |
Neg logit contributions | |
ried | -0.541 |
ck | -0.514 |
ases | -0.502 |
aunts | -0.475 |
beh | -0.461 |
Token | a |
Token ID | 257 |
Feature activation | -4.57e+00 |
Pos logit contributions | |
brand | +0.114 |
� | +0.109 |
truly | +0.102 |
Not | +0.100 |
Imagine | +0.100 |
Neg logit contributions | |
ried | -0.155 |
ck | -0.143 |
ases | -0.139 |
aunts | -0.130 |
ped | -0.126 |
Token | popular |
Token ID | 2968 |
Feature activation | -2.13e+00 |
Pos logit contributions | |
� | +0.120 |
Not | +0.118 |
Eventually | +0.113 |
Finally | +0.113 |
Imagine | +0.113 |
Neg logit contributions | |
ried | -0.127 |
ck | -0.123 |
ies | -0.117 |
leave | -0.115 |
share | -0.113 |
Token | toy |
Token ID | 13373 |
Feature activation | -4.55e+00 |
Pos logit contributions | |
brand | +0.061 |
� | +0.055 |
truly | +0.055 |
sky | +0.053 |
breathtaking | +0.052 |
Neg logit contributions | |
ried | -0.075 |
ck | -0.070 |
ases | -0.068 |
aunts | -0.063 |
beh | -0.061 |
Token | , |
Token ID | 11 |
Feature activation | -2.92e+00 |
Pos logit contributions | |
brand | +0.152 |
� | +0.139 |
truly | +0.139 |
sky | +0.138 |
Not | +0.129 |
Neg logit contributions | |
ried | -0.144 |
ck | -0.136 |
ases | -0.127 |
aunts | -0.121 |
beh | -0.119 |
Token | a |
Token ID | 257 |
Feature activation | -4.64e+00 |
Pos logit contributions | |
brand | +0.086 |
� | +0.080 |
truly | +0.077 |
Imagine | +0.075 |
sky | +0.075 |
Neg logit contributions | |
ried | -0.096 |
ck | -0.090 |
ases | -0.086 |
aunts | -0.080 |
race | -0.078 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.74e+01 |
Pos logit contributions | |
� | +0.121 |
brand | +0.119 |
Not | +0.118 |
truly | +0.114 |
Imagine | +0.114 |
Neg logit contributions | |
ried | -0.142 |
ck | -0.134 |
ases | -0.124 |
ies | -0.122 |
steer | -0.120 |
Token | new |
Token ID | 649 |
Feature activation | -3.68e+00 |
Pos logit contributions | |
brand | +0.414 |
� | +0.413 |
Not | +0.390 |
Imagine | +0.390 |
truly | +0.385 |
Neg logit contributions | |
ried | -0.631 |
ck | -0.596 |
ases | -0.574 |
aunts | -0.545 |
ies | -0.535 |
Token | doll |
Token ID | 3654 |
Feature activation | -3.95e-01 |
Pos logit contributions | |
brand | +0.114 |
� | +0.103 |
truly | +0.103 |
sky | +0.100 |
breathtaking | +0.098 |
Neg logit contributions | |
ried | -0.120 |
ck | -0.111 |
ases | -0.108 |
aunts | -0.100 |
beh | -0.096 |
Token | . |
Token ID | 13 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
brand | +0.014 |
truly | +0.013 |
� | +0.013 |
sky | +0.012 |
Not | +0.012 |
Neg logit contributions | |
ried | -0.012 |
ck | -0.011 |
ases | -0.010 |
aunts | -0.010 |
ucks | -0.009 |
Token |
|
Token ID | 198 |
Feature activation | -4.92e+00 |
Pos logit contributions | |
ried | +0.043 |
ck | +0.037 |
ases | +0.036 |
ucks | +0.036 |
aunts | +0.035 |
Neg logit contributions | |
brand | -0.043 |
truly | -0.038 |
� | -0.037 |
breathtaking | -0.037 |
Imagine | -0.037 |
Token |
|
Token ID | 198 |
Feature activation | -5.00e+00 |
Pos logit contributions | |
brand | +0.119 |
� | +0.105 |
truly | +0.102 |
sky | +0.098 |
Not | +0.097 |
Neg logit contributions | |
ried | -0.191 |
ck | -0.182 |
ases | -0.174 |
aunts | -0.166 |
race | -0.161 |
Token | confidently |
Token ID | 37891 |
Feature activation | +4.73e+00 |
Pos logit contributions | |
ck | +0.185 |
ases | +0.174 |
ried | +0.170 |
beh | +0.163 |
aunts | +0.155 |
Neg logit contributions | |
brand | -0.235 |
sky | -0.220 |
truly | -0.212 |
so | -0.212 |
most | -0.203 |
Token | back |
Token ID | 736 |
Feature activation | +3.08e+00 |
Pos logit contributions | |
ried | +0.126 |
ck | +0.120 |
ases | +0.115 |
aunts | +0.109 |
ucks | +0.103 |
Neg logit contributions | |
brand | -0.168 |
truly | -0.156 |
sky | -0.155 |
� | -0.151 |
Not | -0.148 |
Token | home |
Token ID | 1363 |
Feature activation | -1.88e+00 |
Pos logit contributions | |
ck | +0.090 |
ried | +0.087 |
ases | +0.080 |
beh | +0.077 |
aunts | +0.077 |
Neg logit contributions | |
brand | -0.103 |
sky | -0.097 |
truly | -0.094 |
� | -0.090 |
Once | -0.088 |
Token | with |
Token ID | 351 |
Feature activation | +3.38e+00 |
Pos logit contributions | |
brand | +0.066 |
truly | +0.061 |
� | +0.060 |
sky | +0.060 |
Not | +0.058 |
Neg logit contributions | |
ried | -0.052 |
ck | -0.050 |
ases | -0.049 |
aunts | -0.044 |
beh | -0.043 |
Token | a |
Token ID | 257 |
Feature activation | -4.64e+00 |
Pos logit contributions | |
ried | +0.112 |
ck | +0.100 |
ases | +0.098 |
aunts | +0.093 |
beh | +0.089 |
Neg logit contributions | |
brand | -0.110 |
truly | -0.099 |
� | -0.096 |
sky | -0.094 |
breathtaking | -0.093 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.74e+01 |
Pos logit contributions | |
Not | +0.143 |
� | +0.141 |
brand | +0.136 |
Imagine | +0.135 |
Too | +0.135 |
Neg logit contributions | |
ried | -0.115 |
ck | -0.107 |
steer | -0.103 |
leave | -0.099 |
ies | -0.099 |
Token | new |
Token ID | 649 |
Feature activation | -3.74e+00 |
Pos logit contributions | |
brand | +0.427 |
� | +0.401 |
truly | +0.389 |
Imagine | +0.384 |
Not | +0.382 |
Neg logit contributions | |
ried | -0.639 |
ck | -0.601 |
ases | -0.585 |
aunts | -0.553 |
ucks | -0.538 |
Token | haircut |
Token ID | 45548 |
Feature activation | -5.67e-01 |
Pos logit contributions | |
brand | +0.127 |
� | +0.117 |
truly | +0.117 |
Not | +0.114 |
Imagine | +0.113 |
Neg logit contributions | |
ried | -0.106 |
ck | -0.097 |
ases | -0.094 |
aunts | -0.086 |
race | -0.083 |
Token | . |
Token ID | 13 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
brand | +0.023 |
� | +0.021 |
truly | +0.021 |
sky | +0.021 |
Not | +0.020 |
Neg logit contributions | |
ried | -0.013 |
ck | -0.013 |
ases | -0.012 |
aunts | -0.011 |
beh | -0.010 |
Token | He |
Token ID | 679 |
Feature activation | -4.80e-01 |
Pos logit contributions | |
ried | +0.045 |
ases | +0.041 |
ck | +0.040 |
aunts | +0.038 |
race | +0.037 |
Neg logit contributions | |
brand | -0.041 |
truly | -0.037 |
� | -0.036 |
breathtaking | -0.036 |
Imagine | -0.036 |
Token | was |
Token ID | 373 |
Feature activation | -4.37e+00 |
Pos logit contributions | |
brand | +0.014 |
truly | +0.013 |
sky | +0.012 |
Imagine | +0.012 |
� | +0.011 |
Neg logit contributions | |
ried | -0.017 |
ases | -0.016 |
ck | -0.015 |
aunts | -0.015 |
ucks | -0.014 |
Token | he |
Token ID | 339 |
Feature activation | -9.57e-01 |
Pos logit contributions | |
brand | +0.318 |
� | +0.299 |
truly | +0.294 |
Not | +0.291 |
Imagine | +0.289 |
Neg logit contributions | |
ried | -0.251 |
ck | -0.217 |
ases | -0.207 |
ie | -0.191 |
aunts | -0.190 |
Token | discovered |
Token ID | 5071 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
brand | +0.028 |
truly | +0.025 |
Imagine | +0.024 |
� | +0.024 |
sky | +0.024 |
Neg logit contributions | |
ried | -0.033 |
ases | -0.031 |
ck | -0.031 |
beh | -0.029 |
aunts | -0.028 |
Token | it |
Token ID | 340 |
Feature activation | -9.21e+00 |
Pos logit contributions | |
ried | +0.035 |
ck | +0.033 |
ases | +0.032 |
aunts | +0.029 |
ped | +0.029 |
Neg logit contributions | |
brand | -0.031 |
� | -0.030 |
truly | -0.028 |
Imagine | -0.028 |
Not | -0.028 |
Token | was |
Token ID | 373 |
Feature activation | -4.40e+00 |
Pos logit contributions | |
brand | +0.298 |
� | +0.278 |
truly | +0.269 |
sky | +0.265 |
Not | +0.265 |
Neg logit contributions | |
ried | -0.278 |
ck | -0.256 |
ases | -0.244 |
aunts | -0.224 |
race | -0.217 |
Token | a |
Token ID | 257 |
Feature activation | -4.57e+00 |
Pos logit contributions | |
brand | +0.130 |
� | +0.118 |
truly | +0.117 |
sky | +0.113 |
Imagine | +0.112 |
Neg logit contributions | |
ried | -0.149 |
ck | -0.138 |
ases | -0.134 |
aunts | -0.125 |
race | -0.120 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.73e+01 |
Pos logit contributions | |
� | +0.122 |
Not | +0.120 |
Eventually | +0.114 |
Finally | +0.110 |
Too | +0.110 |
Neg logit contributions | |
ried | -0.127 |
ck | -0.123 |
share | -0.118 |
steer | -0.115 |
leave | -0.115 |
Token | new |
Token ID | 649 |
Feature activation | -3.65e+00 |
Pos logit contributions | |
brand | +0.407 |
� | +0.402 |
Not | +0.381 |
Imagine | +0.379 |
truly | +0.379 |
Neg logit contributions | |
ried | -0.643 |
ck | -0.606 |
ases | -0.581 |
aunts | -0.553 |
ies | -0.542 |
Token | bike |
Token ID | 7161 |
Feature activation | -9.41e-01 |
Pos logit contributions | |
brand | +0.115 |
� | +0.109 |
truly | +0.107 |
Not | +0.105 |
Imagine | +0.103 |
Neg logit contributions | |
ried | -0.108 |
ck | -0.099 |
ases | -0.096 |
aunts | -0.090 |
ies | -0.089 |
Token | ! |
Token ID | 0 |
Feature activation | +7.12e-01 |
Pos logit contributions | |
brand | +0.033 |
truly | +0.030 |
� | +0.029 |
sky | +0.029 |
breathtaking | +0.028 |
Neg logit contributions | |
ried | -0.028 |
ck | -0.025 |
ases | -0.025 |
aunts | -0.023 |
ucks | -0.022 |
Token |
|
Token ID | 198 |
Feature activation | -4.99e+00 |
Pos logit contributions | |
ried | +0.026 |
ck | +0.024 |
ases | +0.023 |
aunts | +0.022 |
race | +0.021 |
Neg logit contributions | |
brand | -0.019 |
truly | -0.017 |
� | -0.017 |
breathtaking | -0.016 |
sky | -0.016 |
Token |
|
Token ID | 198 |
Feature activation | -5.06e+00 |
Pos logit contributions | |
brand | +0.112 |
� | +0.098 |
truly | +0.097 |
sky | +0.093 |
breathtaking | +0.092 |
Neg logit contributions | |
ried | -0.202 |
ck | -0.191 |
ases | -0.187 |
aunts | -0.176 |
race | -0.172 |
Token | was |
Token ID | 373 |
Feature activation | -4.30e+00 |
Pos logit contributions | |
brand | +0.222 |
� | +0.196 |
truly | +0.194 |
sky | +0.186 |
Not | +0.185 |
Neg logit contributions | |
ried | -0.358 |
ck | -0.336 |
ases | -0.327 |
aunts | -0.308 |
race | -0.298 |
Token | dry |
Token ID | 5894 |
Feature activation | +6.81e-01 |
Pos logit contributions | |
brand | +0.128 |
� | +0.117 |
truly | +0.116 |
Not | +0.112 |
sky | +0.112 |
Neg logit contributions | |
ried | -0.143 |
ck | -0.133 |
ases | -0.128 |
aunts | -0.118 |
race | -0.114 |
Token | , |
Token ID | 11 |
Feature activation | -3.01e+00 |
Pos logit contributions | |
ried | +0.022 |
ck | +0.021 |
ases | +0.021 |
aunts | +0.019 |
beh | +0.019 |
Neg logit contributions | |
brand | -0.021 |
� | -0.019 |
truly | -0.019 |
sky | -0.018 |
breathtaking | -0.018 |
Token | it |
Token ID | 340 |
Feature activation | -9.16e+00 |
Pos logit contributions | |
brand | +0.109 |
� | +0.101 |
truly | +0.100 |
Imagine | +0.097 |
breathtaking | +0.097 |
Neg logit contributions | |
ried | -0.080 |
ck | -0.073 |
ases | -0.071 |
aunts | -0.065 |
race | -0.062 |
Token | looked |
Token ID | 3114 |
Feature activation | -6.35e+00 |
Pos logit contributions | |
brand | +0.269 |
truly | +0.245 |
� | +0.238 |
sky | +0.235 |
Imagine | +0.230 |
Neg logit contributions | |
ried | -0.310 |
ases | -0.289 |
ck | -0.287 |
aunts | -0.265 |
race | -0.258 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.73e+01 |
Pos logit contributions | |
� | +0.150 |
Not | +0.131 |
brand | +0.129 |
Imagine | +0.128 |
sky | +0.126 |
Neg logit contributions | |
ried | -0.246 |
ck | -0.231 |
ases | -0.214 |
leave | -0.206 |
ies | -0.206 |
Token | - |
Token ID | 12 |
Feature activation | +2.56e-01 |
Pos logit contributions | |
� | +0.478 |
Imagine | +0.452 |
Not | +0.451 |
sky | +0.431 |
fact | +0.430 |
Neg logit contributions | |
ried | -0.554 |
ck | -0.540 |
ie | -0.497 |
ases | -0.489 |
ies | -0.480 |
Token | new |
Token ID | 3605 |
Feature activation | -2.70e+00 |
Pos logit contributions | |
ried | +0.006 |
ases | +0.006 |
ck | +0.006 |
aunts | +0.005 |
pa | +0.005 |
Neg logit contributions | |
brand | -0.010 |
truly | -0.009 |
� | -0.008 |
breathtaking | -0.008 |
Imagine | -0.008 |
Token | again |
Token ID | 757 |
Feature activation | -2.17e+00 |
Pos logit contributions | |
brand | +0.098 |
truly | +0.089 |
� | +0.089 |
sky | +0.087 |
breathtaking | +0.086 |
Neg logit contributions | |
ried | -0.074 |
ck | -0.068 |
ases | -0.066 |
aunts | -0.060 |
race | -0.058 |
Token | ! |
Token ID | 0 |
Feature activation | +6.42e-01 |
Pos logit contributions | |
brand | +0.079 |
truly | +0.073 |
� | +0.072 |
sky | +0.071 |
breathtaking | +0.070 |
Neg logit contributions | |
ried | -0.059 |
ck | -0.054 |
ases | -0.053 |
aunts | -0.048 |
beh | -0.046 |
Token | L |
Token ID | 406 |
Feature activation | +6.74e+00 |
Pos logit contributions | |
ried | +0.019 |
ases | +0.018 |
ck | +0.017 |
aunts | +0.016 |
ucks | +0.016 |
Neg logit contributions | |
brand | -0.020 |
breathtaking | -0.018 |
truly | -0.018 |
� | -0.018 |
Imagine | -0.018 |
Token | what |
Token ID | 644 |
Feature activation | -6.16e+00 |
Pos logit contributions | |
ried | +0.081 |
ck | +0.072 |
ases | +0.070 |
aunts | +0.065 |
beh | +0.062 |
Neg logit contributions | |
brand | -0.114 |
truly | -0.104 |
sky | -0.102 |
� | -0.102 |
breathtaking | -0.101 |
Token | it |
Token ID | 340 |
Feature activation | -9.30e+00 |
Pos logit contributions | |
brand | +0.190 |
� | +0.172 |
truly | +0.171 |
sky | +0.166 |
Imagine | +0.164 |
Neg logit contributions | |
ried | -0.200 |
ck | -0.185 |
ases | -0.181 |
aunts | -0.167 |
race | -0.161 |
Token | was |
Token ID | 373 |
Feature activation | -4.50e+00 |
Pos logit contributions | |
brand | +0.233 |
� | +0.209 |
truly | +0.196 |
Not | +0.196 |
breathtaking | +0.193 |
Neg logit contributions | |
ried | -0.357 |
ck | -0.332 |
ases | -0.318 |
aunts | -0.301 |
ies | -0.293 |
Token | - |
Token ID | 532 |
Feature activation | -3.52e+00 |
Pos logit contributions | |
brand | +0.167 |
� | +0.159 |
Not | +0.152 |
Imagine | +0.151 |
truly | +0.151 |
Neg logit contributions | |
ried | -0.113 |
ck | -0.102 |
ases | -0.096 |
aunts | -0.087 |
ies | -0.086 |
Token | a |
Token ID | 257 |
Feature activation | -4.64e+00 |
Pos logit contributions | |
brand | +0.106 |
� | +0.100 |
truly | +0.096 |
Imagine | +0.094 |
Not | +0.093 |
Neg logit contributions | |
ried | -0.111 |
ck | -0.103 |
ases | -0.099 |
aunts | -0.092 |
race | -0.091 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.73e+01 |
Pos logit contributions | |
� | +0.126 |
Not | +0.124 |
brand | +0.121 |
Too | +0.116 |
Cookies | +0.115 |
Neg logit contributions | |
ried | -0.139 |
ck | -0.131 |
ies | -0.119 |
leave | -0.118 |
ases | -0.117 |
Token | new |
Token ID | 649 |
Feature activation | -3.61e+00 |
Pos logit contributions | |
brand | +0.408 |
� | +0.403 |
Not | +0.378 |
truly | +0.376 |
Imagine | +0.375 |
Neg logit contributions | |
ried | -0.645 |
ck | -0.607 |
ases | -0.585 |
aunts | -0.555 |
ies | -0.545 |
Token | swing |
Token ID | 9628 |
Feature activation | +3.12e+00 |
Pos logit contributions | |
brand | +0.125 |
� | +0.116 |
truly | +0.114 |
Not | +0.111 |
Imagine | +0.110 |
Neg logit contributions | |
ried | -0.101 |
ck | -0.092 |
ases | -0.090 |
aunts | -0.082 |
race | -0.079 |
Token | ! |
Token ID | 0 |
Feature activation | +6.40e-01 |
Pos logit contributions | |
ried | +0.096 |
ck | +0.089 |
ases | +0.087 |
aunts | +0.081 |
beh | +0.077 |
Neg logit contributions | |
brand | -0.101 |
� | -0.092 |
truly | -0.092 |
sky | -0.090 |
Imagine | -0.088 |
Token | She |
Token ID | 1375 |
Feature activation | +1.20e-01 |
Pos logit contributions | |
ried | +0.020 |
ases | +0.018 |
ck | +0.018 |
aunts | +0.017 |
race | +0.017 |
Neg logit contributions | |
brand | -0.020 |
truly | -0.018 |
breathtaking | -0.017 |
Imagine | -0.017 |
� | -0.017 |
Token | thanked |
Token ID | 26280 |
Feature activation | +8.52e+00 |
Pos logit contributions | |
ried | +0.004 |
ases | +0.003 |
ck | +0.003 |
aunts | +0.003 |
beh | +0.003 |
Neg logit contributions | |
brand | -0.004 |
truly | -0.004 |
� | -0.004 |
sky | -0.004 |
Imagine | -0.004 |
Token | see |
Token ID | 766 |
Feature activation | +2.96e+00 |
Pos logit contributions | |
ried | +0.178 |
ases | +0.150 |
ck | +0.147 |
beh | +0.143 |
ucks | +0.143 |
Neg logit contributions | |
brand | -0.190 |
truly | -0.184 |
Not | -0.177 |
� | -0.172 |
definitely | -0.171 |
Token | that |
Token ID | 326 |
Feature activation | -3.70e+00 |
Pos logit contributions | |
ried | +0.107 |
ck | +0.096 |
ases | +0.095 |
aunts | +0.089 |
beh | +0.087 |
Neg logit contributions | |
brand | -0.086 |
truly | -0.076 |
breathtaking | -0.071 |
� | -0.071 |
sky | -0.071 |
Token | it |
Token ID | 340 |
Feature activation | -9.08e+00 |
Pos logit contributions | |
brand | +0.109 |
truly | +0.098 |
sky | +0.095 |
� | +0.094 |
breathtaking | +0.092 |
Neg logit contributions | |
ried | -0.130 |
ck | -0.120 |
ases | -0.118 |
aunts | -0.109 |
beh | -0.106 |
Token | was |
Token ID | 373 |
Feature activation | -4.12e+00 |
Pos logit contributions | |
brand | +0.216 |
� | +0.193 |
truly | +0.186 |
sky | +0.181 |
Not | +0.180 |
Neg logit contributions | |
ried | -0.358 |
ck | -0.335 |
ases | -0.325 |
aunts | -0.305 |
race | -0.297 |
Token | a |
Token ID | 257 |
Feature activation | -4.31e+00 |
Pos logit contributions | |
brand | +0.116 |
truly | +0.103 |
breathtaking | +0.096 |
sky | +0.094 |
� | +0.092 |
Neg logit contributions | |
ried | -0.153 |
ases | -0.145 |
ck | -0.141 |
aunts | -0.135 |
ucks | -0.132 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.73e+01 |
Pos logit contributions | |
� | +0.115 |
Not | +0.110 |
Eventually | +0.108 |
brand | +0.107 |
Too | +0.105 |
Neg logit contributions | |
ried | -0.118 |
ck | -0.118 |
share | -0.115 |
leave | -0.110 |
steer | -0.110 |
Token | new |
Token ID | 649 |
Feature activation | -3.51e+00 |
Pos logit contributions | |
brand | +0.408 |
� | +0.391 |
truly | +0.373 |
Imagine | +0.369 |
Not | +0.369 |
Neg logit contributions | |
ried | -0.649 |
ck | -0.612 |
ases | -0.592 |
aunts | -0.561 |
ucks | -0.545 |
Token | puzzle |
Token ID | 15027 |
Feature activation | -8.32e+00 |
Pos logit contributions | |
brand | +0.097 |
� | +0.089 |
truly | +0.087 |
Not | +0.085 |
breathtaking | +0.085 |
Neg logit contributions | |
ried | -0.118 |
ck | -0.110 |
ases | -0.108 |
aunts | -0.101 |
race | -0.099 |
Token | ! |
Token ID | 0 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
brand | +0.334 |
truly | +0.312 |
sky | +0.302 |
� | +0.302 |
breathtaking | +0.293 |
Neg logit contributions | |
ried | -0.202 |
ck | -0.182 |
ases | -0.174 |
aunts | -0.162 |
ucks | -0.155 |
Token | He |
Token ID | 679 |
Feature activation | -2.38e-01 |
Pos logit contributions | |
ried | +0.035 |
ases | +0.033 |
aunts | +0.032 |
his | +0.032 |
him | +0.031 |
Neg logit contributions | |
brand | -0.045 |
truly | -0.040 |
breathtaking | -0.038 |
� | -0.037 |
fantastic | -0.037 |
Token | put |
Token ID | 1234 |
Feature activation | +5.51e+00 |
Pos logit contributions | |
brand | +0.008 |
truly | +0.007 |
� | +0.007 |
Imagine | +0.007 |
sky | +0.007 |
Neg logit contributions | |
ried | -0.007 |
ck | -0.007 |
ases | -0.007 |
aunts | -0.006 |
beh | -0.006 |
Token | the |
Token ID | 262 |
Feature activation | -8.24e-01 |
Pos logit contributions | |
ried | +0.161 |
ases | +0.148 |
ck | +0.144 |
beh | +0.137 |
aunts | +0.134 |
Neg logit contributions | |
brand | -0.231 |
sky | -0.211 |
truly | -0.208 |
breathtaking | -0.207 |
� | -0.204 |
Token | sack |
Token ID | 23704 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
� | +0.031 |
brand | +0.031 |
truly | +0.030 |
Not | +0.030 |
Imagine | +0.030 |
Neg logit contributions | |
ried | -0.018 |
leave | -0.014 |
ck | -0.014 |
ases | -0.014 |
ies | -0.013 |
Token | and |
Token ID | 290 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
brand | +0.077 |
� | +0.071 |
truly | +0.070 |
Not | +0.069 |
breathtaking | +0.068 |
Neg logit contributions | |
ried | -0.056 |
ck | -0.048 |
ases | -0.047 |
aunts | -0.043 |
race | -0.043 |
Token | found |
Token ID | 1043 |
Feature activation | +2.98e+00 |
Pos logit contributions | |
brand | +0.044 |
truly | +0.040 |
Imagine | +0.039 |
sky | +0.038 |
breathtaking | +0.037 |
Neg logit contributions | |
ases | -0.050 |
ried | -0.045 |
beh | -0.043 |
ck | -0.043 |
aunts | -0.043 |
Token | a |
Token ID | 257 |
Feature activation | -4.38e+00 |
Pos logit contributions | |
ried | +0.104 |
ck | +0.097 |
ases | +0.095 |
aunts | +0.089 |
beh | +0.086 |
Neg logit contributions | |
brand | -0.085 |
truly | -0.076 |
� | -0.075 |
sky | -0.073 |
breathtaking | -0.072 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.73e+01 |
Pos logit contributions | |
� | +0.122 |
Not | +0.121 |
Eventually | +0.112 |
Very | +0.112 |
Cookies | +0.110 |
Neg logit contributions | |
ried | -0.124 |
ck | -0.114 |
ies | -0.112 |
share | -0.110 |
steer | -0.108 |
Token | new |
Token ID | 649 |
Feature activation | -3.69e+00 |
Pos logit contributions | |
brand | +0.410 |
� | +0.393 |
truly | +0.374 |
Not | +0.372 |
Imagine | +0.369 |
Neg logit contributions | |
ried | -0.649 |
ck | -0.612 |
ases | -0.591 |
aunts | -0.562 |
ies | -0.550 |
Token | bright |
Token ID | 6016 |
Feature activation | -4.07e+00 |
Pos logit contributions | |
brand | +0.118 |
� | +0.108 |
truly | +0.107 |
Not | +0.104 |
sky | +0.104 |
Neg logit contributions | |
ried | -0.113 |
ck | -0.104 |
ases | -0.101 |
aunts | -0.093 |
race | -0.090 |
Token | yellow |
Token ID | 7872 |
Feature activation | -2.95e+00 |
Pos logit contributions | |
brand | +0.119 |
� | +0.115 |
Not | +0.114 |
truly | +0.112 |
Finally | +0.111 |
Neg logit contributions | |
ried | -0.123 |
ck | -0.115 |
ases | -0.109 |
ies | -0.106 |
ie | -0.104 |
Token | ball |
Token ID | 2613 |
Feature activation | -3.43e+00 |
Pos logit contributions | |
brand | +0.092 |
� | +0.084 |
truly | +0.084 |
Not | +0.081 |
Imagine | +0.080 |
Neg logit contributions | |
ried | -0.093 |
ck | -0.086 |
ases | -0.084 |
aunts | -0.078 |
race | -0.075 |
Token | ! |
Token ID | 0 |
Feature activation | +6.87e-01 |
Pos logit contributions | |
brand | +0.130 |
sky | +0.122 |
� | +0.119 |
truly | +0.119 |
breathtaking | +0.111 |
Neg logit contributions | |
ried | -0.086 |
ck | -0.083 |
ases | -0.079 |
aunts | -0.076 |
ucks | -0.074 |
Token | joy |
Token ID | 8716 |
Feature activation | -3.12e+00 |
Pos logit contributions | |
ried | +0.113 |
ck | +0.105 |
ases | +0.103 |
aunts | +0.094 |
race | +0.091 |
Neg logit contributions | |
brand | -0.111 |
� | -0.102 |
truly | -0.101 |
Imagine | -0.098 |
sky | -0.098 |
Token | - |
Token ID | 532 |
Feature activation | -3.40e+00 |
Pos logit contributions | |
brand | +0.123 |
truly | +0.112 |
sky | +0.112 |
breathtaking | +0.107 |
� | +0.105 |
Neg logit contributions | |
ried | -0.077 |
ases | -0.071 |
ck | -0.071 |
aunts | -0.064 |
ucks | -0.061 |
Token | it |
Token ID | 340 |
Feature activation | -9.06e+00 |
Pos logit contributions | |
brand | +0.098 |
truly | +0.087 |
� | +0.085 |
sky | +0.084 |
Not | +0.083 |
Neg logit contributions | |
ried | -0.122 |
ck | -0.112 |
ases | -0.110 |
aunts | -0.103 |
beh | -0.099 |
Token | was |
Token ID | 373 |
Feature activation | -4.14e+00 |
Pos logit contributions | |
brand | +0.214 |
� | +0.191 |
truly | +0.187 |
sky | +0.180 |
Not | +0.179 |
Neg logit contributions | |
ried | -0.356 |
ck | -0.336 |
ases | -0.326 |
aunts | -0.307 |
race | -0.298 |
Token | a |
Token ID | 257 |
Feature activation | -4.54e+00 |
Pos logit contributions | |
brand | +0.127 |
truly | +0.115 |
� | +0.112 |
sky | +0.111 |
breathtaking | +0.110 |
Neg logit contributions | |
ried | -0.137 |
ck | -0.126 |
ases | -0.124 |
aunts | -0.115 |
race | -0.111 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.73e+01 |
Pos logit contributions | |
� | +0.125 |
brand | +0.123 |
Not | +0.122 |
truly | +0.119 |
Finally | +0.119 |
Neg logit contributions | |
ried | -0.128 |
ck | -0.126 |
ases | -0.114 |
steer | -0.112 |
ies | -0.112 |
Token | new |
Token ID | 649 |
Feature activation | -3.62e+00 |
Pos logit contributions | |
� | +0.423 |
brand | +0.412 |
Imagine | +0.403 |
Not | +0.402 |
sky | +0.391 |
Neg logit contributions | |
ried | -0.611 |
ck | -0.585 |
ases | -0.552 |
aunts | -0.529 |
ies | -0.527 |
Token | doll |
Token ID | 3654 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
brand | +0.113 |
� | +0.106 |
truly | +0.104 |
Not | +0.102 |
Imagine | +0.101 |
Neg logit contributions | |
ried | -0.109 |
ck | -0.101 |
ases | -0.099 |
aunts | -0.091 |
ies | -0.089 |
Token | ! |
Token ID | 0 |
Feature activation | +7.16e-01 |
Pos logit contributions | |
brand | +0.008 |
truly | +0.007 |
� | +0.007 |
sky | +0.007 |
breathtaking | +0.007 |
Neg logit contributions | |
ried | -0.009 |
ases | -0.009 |
ck | -0.008 |
ucks | -0.008 |
aunts | -0.008 |
Token | All |
Token ID | 1439 |
Feature activation | -5.38e+00 |
Pos logit contributions | |
ried | +0.023 |
ases | +0.021 |
ck | +0.021 |
aunts | +0.020 |
race | +0.019 |
Neg logit contributions | |
brand | -0.021 |
truly | -0.019 |
breathtaking | -0.019 |
� | -0.019 |
Imagine | -0.018 |
Token | of |
Token ID | 286 |
Feature activation | +5.08e+00 |
Pos logit contributions | |
brand | +0.202 |
truly | +0.188 |
sky | +0.185 |
breathtaking | +0.180 |
Imagine | +0.179 |
Neg logit contributions | |
ried | -0.139 |
ck | -0.125 |
ases | -0.121 |
aunts | -0.112 |
ucks | -0.106 |
Token | happy |
Token ID | 3772 |
Feature activation | -6.61e+00 |
Pos logit contributions | |
brand | +0.355 |
� | +0.349 |
sky | +0.329 |
Not | +0.321 |
Imagine | +0.321 |
Neg logit contributions | |
ried | -0.403 |
ck | -0.375 |
ases | -0.350 |
aunts | -0.332 |
share | -0.321 |
Token | ! |
Token ID | 0 |
Feature activation | +9.03e-01 |
Pos logit contributions | |
brand | +0.247 |
truly | +0.228 |
� | +0.228 |
sky | +0.222 |
Not | +0.220 |
Neg logit contributions | |
ried | -0.173 |
ck | -0.158 |
ases | -0.152 |
aunts | -0.138 |
race | -0.130 |
Token | Inside |
Token ID | 14384 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
ried | +0.028 |
ases | +0.025 |
ck | +0.024 |
race | +0.023 |
aunts | +0.023 |
Neg logit contributions | |
brand | -0.029 |
truly | -0.026 |
breathtaking | -0.026 |
� | -0.025 |
Imagine | -0.025 |
Token | was |
Token ID | 373 |
Feature activation | -4.32e+00 |
Pos logit contributions | |
brand | +0.040 |
� | +0.035 |
Not | +0.035 |
truly | +0.035 |
Imagine | +0.035 |
Neg logit contributions | |
ried | -0.035 |
ck | -0.032 |
ases | -0.031 |
ie | -0.029 |
aunts | -0.028 |
Token | a |
Token ID | 257 |
Feature activation | -4.59e+00 |
Pos logit contributions | |
brand | +0.137 |
� | +0.133 |
Imagine | +0.127 |
Not | +0.124 |
truly | +0.124 |
Neg logit contributions | |
ried | -0.127 |
ck | -0.115 |
ases | -0.111 |
aunts | -0.101 |
ped | -0.099 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.73e+01 |
Pos logit contributions | |
� | +0.121 |
Not | +0.117 |
Cookies | +0.112 |
Too | +0.111 |
pill | +0.108 |
Neg logit contributions | |
share | -0.120 |
ried | -0.120 |
ck | -0.119 |
pet | -0.117 |
take | -0.117 |
Token | new |
Token ID | 649 |
Feature activation | -3.63e+00 |
Pos logit contributions | |
brand | +0.415 |
� | +0.413 |
Imagine | +0.394 |
Not | +0.393 |
truly | +0.387 |
Neg logit contributions | |
ried | -0.626 |
ck | -0.589 |
ases | -0.567 |
aunts | -0.538 |
ies | -0.530 |
Token | toy |
Token ID | 13373 |
Feature activation | -4.53e+00 |
Pos logit contributions | |
brand | +0.111 |
� | +0.108 |
Not | +0.105 |
truly | +0.103 |
Finally | +0.102 |
Neg logit contributions | |
ried | -0.103 |
ck | -0.094 |
ases | -0.091 |
ies | -0.088 |
pa | -0.085 |
Token | that |
Token ID | 326 |
Feature activation | -3.95e+00 |
Pos logit contributions | |
brand | +0.142 |
truly | +0.127 |
� | +0.127 |
sky | +0.126 |
Not | +0.117 |
Neg logit contributions | |
ried | -0.155 |
ck | -0.146 |
ases | -0.137 |
aunts | -0.131 |
beh | -0.130 |
Token | she |
Token ID | 673 |
Feature activation | -5.21e-01 |
Pos logit contributions | |
brand | +0.112 |
� | +0.103 |
truly | +0.098 |
Not | +0.097 |
Imagine | +0.097 |
Neg logit contributions | |
ried | -0.136 |
ck | -0.125 |
ases | -0.121 |
aunts | -0.113 |
race | -0.110 |
Token | could |
Token ID | 714 |
Feature activation | -3.58e+00 |
Pos logit contributions | |
brand | +0.013 |
truly | +0.012 |
Imagine | +0.010 |
definitely | +0.010 |
sky | +0.010 |
Neg logit contributions | |
ried | -0.020 |
ases | -0.020 |
ck | -0.019 |
ucks | -0.019 |
aunts | -0.019 |
Token | day |
Token ID | 1110 |
Feature activation | -8.83e+00 |
Pos logit contributions | |
brand | +0.062 |
� | +0.054 |
truly | +0.053 |
sky | +0.049 |
breathtaking | +0.049 |
Neg logit contributions | |
ried | -0.133 |
ck | -0.126 |
ases | -0.123 |
aunts | -0.117 |
race | -0.113 |
Token | , |
Token ID | 11 |
Feature activation | -2.48e+00 |
Pos logit contributions | |
brand | +0.256 |
� | +0.234 |
truly | +0.229 |
sky | +0.225 |
Not | +0.224 |
Neg logit contributions | |
ried | -0.307 |
ck | -0.283 |
ases | -0.278 |
aunts | -0.256 |
beh | -0.249 |
Token | she |
Token ID | 673 |
Feature activation | -1.24e-01 |
Pos logit contributions | |
brand | +0.093 |
truly | +0.085 |
� | +0.085 |
breathtaking | +0.083 |
sky | +0.083 |
Neg logit contributions | |
ried | -0.064 |
ck | -0.058 |
ases | -0.056 |
aunts | -0.051 |
race | -0.049 |
Token | got |
Token ID | 1392 |
Feature activation | -2.64e+00 |
Pos logit contributions | |
brand | +0.003 |
Imagine | +0.003 |
sky | +0.003 |
truly | +0.003 |
Not | +0.003 |
Neg logit contributions | |
ried | -0.004 |
ck | -0.004 |
ases | -0.004 |
beh | -0.004 |
aunts | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | -4.19e+00 |
Pos logit contributions | |
brand | +0.072 |
truly | +0.064 |
sky | +0.061 |
� | +0.061 |
Imagine | +0.059 |
Neg logit contributions | |
ried | -0.096 |
ck | -0.091 |
ases | -0.089 |
aunts | -0.083 |
steer | -0.080 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.73e+01 |
Pos logit contributions | |
� | +0.122 |
Not | +0.118 |
Eventually | +0.116 |
Finally | +0.115 |
Imagine | +0.115 |
Neg logit contributions | |
ried | -0.114 |
ck | -0.110 |
ies | -0.104 |
leave | -0.102 |
ie | -0.100 |
Token | new |
Token ID | 649 |
Feature activation | -3.47e+00 |
Pos logit contributions | |
brand | +0.413 |
� | +0.398 |
truly | +0.382 |
Not | +0.375 |
Imagine | +0.375 |
Neg logit contributions | |
ried | -0.645 |
ck | -0.606 |
ases | -0.589 |
aunts | -0.557 |
ies | -0.544 |
Token | orange |
Token ID | 10912 |
Feature activation | -3.30e+00 |
Pos logit contributions | |
brand | +0.115 |
� | +0.103 |
truly | +0.103 |
sky | +0.101 |
Imagine | +0.099 |
Neg logit contributions | |
ried | -0.107 |
ck | -0.099 |
ases | -0.096 |
aunts | -0.088 |
beh | -0.085 |
Token | dress |
Token ID | 6576 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
brand | +0.104 |
� | +0.096 |
truly | +0.094 |
sky | +0.093 |
Imagine | +0.090 |
Neg logit contributions | |
ried | -0.106 |
ck | -0.097 |
ases | -0.095 |
aunts | -0.087 |
beh | -0.085 |
Token | . |
Token ID | 13 |
Feature activation | +2.14e+00 |
Pos logit contributions | |
brand | +0.075 |
sky | +0.068 |
� | +0.068 |
truly | +0.067 |
breathtaking | +0.063 |
Neg logit contributions | |
ried | -0.057 |
ck | -0.053 |
ases | -0.052 |
aunts | -0.047 |
beh | -0.045 |
Token | She |
Token ID | 1375 |
Feature activation | +5.51e-01 |
Pos logit contributions | |
ried | +0.081 |
ases | +0.075 |
ck | +0.073 |
race | +0.071 |
aunts | +0.071 |
Neg logit contributions | |
brand | -0.051 |
truly | -0.047 |
breathtaking | -0.044 |
� | -0.043 |
Imagine | -0.043 |
Token | store |
Token ID | 3650 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
breathtaking | +0.036 |
brand | +0.035 |
truly | +0.034 |
unlike | +0.033 |
Imagine | +0.033 |
Neg logit contributions | |
ried | -0.021 |
leave | -0.021 |
share | -0.019 |
race | -0.019 |
take | -0.018 |
Token | , |
Token ID | 11 |
Feature activation | -2.74e+00 |
Pos logit contributions | |
brand | +0.038 |
� | +0.034 |
truly | +0.034 |
sky | +0.033 |
Imagine | +0.032 |
Neg logit contributions | |
ried | -0.042 |
ck | -0.040 |
ases | -0.039 |
aunts | -0.036 |
beh | -0.035 |
Token | Sarah |
Token ID | 10490 |
Feature activation | +7.80e+00 |
Pos logit contributions | |
brand | +0.105 |
� | +0.097 |
truly | +0.097 |
sky | +0.094 |
Not | +0.093 |
Neg logit contributions | |
ried | -0.069 |
ck | -0.062 |
ases | -0.060 |
aunts | -0.054 |
beh | -0.051 |
Token | selected |
Token ID | 6163 |
Feature activation | +5.98e-01 |
Pos logit contributions | |
ried | +0.260 |
ck | +0.243 |
ases | +0.240 |
aunts | +0.221 |
beh | +0.214 |
Neg logit contributions | |
brand | -0.232 |
� | -0.210 |
truly | -0.210 |
sky | -0.203 |
Imagine | -0.201 |
Token | a |
Token ID | 257 |
Feature activation | -4.42e+00 |
Pos logit contributions | |
ried | +0.018 |
ck | +0.016 |
ases | +0.016 |
aunts | +0.014 |
race | +0.014 |
Neg logit contributions | |
brand | -0.020 |
� | -0.019 |
truly | -0.018 |
Imagine | -0.018 |
Not | -0.018 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.73e+01 |
Pos logit contributions | |
� | +0.141 |
Not | +0.140 |
Imagine | +0.133 |
Very | +0.131 |
Eventually | +0.131 |
Neg logit contributions | |
ried | -0.094 |
grab | -0.092 |
pet | -0.090 |
take | -0.089 |
share | -0.088 |
Token | new |
Token ID | 649 |
Feature activation | -3.47e+00 |
Pos logit contributions | |
� | +0.470 |
Imagine | +0.457 |
Not | +0.451 |
brand | +0.441 |
truly | +0.435 |
Neg logit contributions | |
ried | -0.553 |
ck | -0.526 |
ases | -0.505 |
ie | -0.495 |
ucks | -0.493 |
Token | brush |
Token ID | 14093 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
brand | +0.132 |
� | +0.126 |
Imagine | +0.124 |
truly | +0.124 |
Not | +0.123 |
Neg logit contributions | |
ried | -0.075 |
ck | -0.067 |
ases | -0.063 |
ies | -0.061 |
race | -0.057 |
Token | . |
Token ID | 13 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
ried | +0.031 |
ck | +0.029 |
ases | +0.028 |
aunts | +0.026 |
beh | +0.024 |
Neg logit contributions | |
brand | -0.039 |
� | -0.036 |
truly | -0.036 |
sky | -0.035 |
breathtaking | -0.035 |
Token | Then |
Token ID | 3244 |
Feature activation | -2.23e+00 |
Pos logit contributions | |
ried | +0.046 |
ases | +0.041 |
race | +0.040 |
aunts | +0.039 |
ucks | +0.038 |
Neg logit contributions | |
brand | -0.045 |
truly | -0.043 |
breathtaking | -0.041 |
fact | -0.040 |
Imagine | -0.039 |
Token | , |
Token ID | 11 |
Feature activation | -2.67e+00 |
Pos logit contributions | |
brand | +0.061 |
truly | +0.054 |
� | +0.054 |
sky | +0.052 |
Not | +0.051 |
Neg logit contributions | |
ried | -0.082 |
ck | -0.077 |
ases | -0.077 |
aunts | -0.071 |
ucks | -0.069 |
Token | store |
Token ID | 3650 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
unlike | +0.039 |
� | +0.038 |
Cookies | +0.038 |
brand | +0.038 |
plets | +0.038 |
Neg logit contributions | |
leave | -0.020 |
take | -0.019 |
share | -0.018 |
his | -0.018 |
her | -0.018 |
Token | , |
Token ID | 11 |
Feature activation | -2.77e+00 |
Pos logit contributions | |
brand | +0.034 |
� | +0.031 |
truly | +0.031 |
sky | +0.031 |
Imagine | +0.030 |
Neg logit contributions | |
ried | -0.040 |
ck | -0.038 |
ases | -0.037 |
aunts | -0.034 |
beh | -0.033 |
Token | Daisy |
Token ID | 40355 |
Feature activation | +6.70e+00 |
Pos logit contributions | |
brand | +0.101 |
� | +0.093 |
truly | +0.092 |
breathtaking | +0.090 |
sky | +0.089 |
Neg logit contributions | |
ried | -0.074 |
ck | -0.066 |
ases | -0.062 |
aunts | -0.058 |
ies | -0.056 |
Token | got |
Token ID | 1392 |
Feature activation | -2.87e+00 |
Pos logit contributions | |
ried | +0.234 |
ck | +0.219 |
ases | +0.215 |
aunts | +0.199 |
race | +0.192 |
Neg logit contributions | |
brand | -0.189 |
� | -0.170 |
truly | -0.170 |
sky | -0.164 |
Imagine | -0.162 |
Token | a |
Token ID | 257 |
Feature activation | -4.45e+00 |
Pos logit contributions | |
brand | +0.088 |
truly | +0.079 |
� | +0.073 |
sky | +0.073 |
much | +0.071 |
Neg logit contributions | |
ried | -0.097 |
ases | -0.094 |
ck | -0.093 |
aunts | -0.086 |
steer | -0.082 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.72e+01 |
Pos logit contributions | |
Not | +0.128 |
Imagine | +0.122 |
sky | +0.122 |
� | +0.121 |
Eventually | +0.121 |
Neg logit contributions | |
ried | -0.113 |
ck | -0.110 |
grab | -0.109 |
take | -0.108 |
share | -0.104 |
Token | new |
Token ID | 649 |
Feature activation | -3.59e+00 |
Pos logit contributions | |
brand | +0.442 |
� | +0.434 |
Imagine | +0.422 |
Not | +0.419 |
truly | +0.414 |
Neg logit contributions | |
ried | -0.594 |
ck | -0.566 |
ases | -0.538 |
aunts | -0.513 |
ie | -0.506 |
Token | , |
Token ID | 11 |
Feature activation | -2.79e+00 |
Pos logit contributions | |
brand | +0.104 |
� | +0.097 |
truly | +0.095 |
Not | +0.095 |
Imagine | +0.093 |
Neg logit contributions | |
ried | -0.115 |
ck | -0.108 |
ases | -0.103 |
aunts | -0.097 |
ies | -0.095 |
Token | fresh |
Token ID | 4713 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
brand | +0.074 |
� | +0.074 |
Imagine | +0.069 |
Not | +0.068 |
truly | +0.068 |
Neg logit contributions | |
ried | -0.094 |
ck | -0.087 |
ases | -0.081 |
leave | -0.077 |
steer | -0.076 |
Token | caul |
Token ID | 46087 |
Feature activation | -5.37e+00 |
Pos logit contributions | |
ried | +0.028 |
ck | +0.027 |
ie | +0.026 |
ies | +0.025 |
share | +0.025 |
Neg logit contributions | |
brand | -0.031 |
truly | -0.030 |
breathtaking | -0.029 |
Not | -0.029 |
Imagine | -0.029 |
Token | iflower |
Token ID | 42642 |
Feature activation | -4.82e+00 |
Pos logit contributions | |
brand | +0.134 |
� | +0.118 |
Not | +0.115 |
truly | +0.114 |
breathtaking | +0.113 |
Neg logit contributions | |
ried | -0.201 |
ck | -0.198 |
ases | -0.184 |
ies | -0.177 |
aunts | -0.176 |
Token | up |
Token ID | 510 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
ried | +0.014 |
ck | +0.013 |
ases | +0.012 |
aunts | +0.010 |
race | +0.010 |
Neg logit contributions | |
brand | -0.033 |
� | -0.031 |
truly | -0.031 |
Not | -0.030 |
Imagine | -0.030 |
Token | as |
Token ID | 355 |
Feature activation | -8.20e+00 |
Pos logit contributions | |
ried | +0.053 |
ck | +0.049 |
ases | +0.049 |
aunts | +0.044 |
beh | +0.043 |
Neg logit contributions | |
brand | -0.062 |
truly | -0.057 |
sky | -0.056 |
� | -0.055 |
breathtaking | -0.055 |
Token | she |
Token ID | 673 |
Feature activation | -1.64e-01 |
Pos logit contributions | |
brand | +0.215 |
truly | +0.196 |
sky | +0.186 |
breathtaking | +0.186 |
� | +0.182 |
Neg logit contributions | |
ried | -0.309 |
ases | -0.288 |
ck | -0.287 |
aunts | -0.268 |
beh | -0.262 |
Token | saw |
Token ID | 2497 |
Feature activation | +3.52e+00 |
Pos logit contributions | |
brand | +0.005 |
� | +0.005 |
truly | +0.005 |
Not | +0.005 |
sky | +0.004 |
Neg logit contributions | |
ried | -0.005 |
ck | -0.005 |
ases | -0.005 |
aunts | -0.004 |
ped | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | -4.20e+00 |
Pos logit contributions | |
ried | +0.115 |
ck | +0.107 |
ases | +0.104 |
aunts | +0.096 |
race | +0.093 |
Neg logit contributions | |
brand | -0.105 |
� | -0.097 |
truly | -0.095 |
Imagine | -0.092 |
Not | -0.092 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.72e+01 |
Pos logit contributions | |
� | +0.110 |
Not | +0.106 |
brand | +0.100 |
Cookies | +0.098 |
Eventually | +0.098 |
Neg logit contributions | |
ried | -0.126 |
ck | -0.124 |
ies | -0.116 |
steer | -0.113 |
ie | -0.110 |
Token | new |
Token ID | 649 |
Feature activation | -3.54e+00 |
Pos logit contributions | |
� | +0.432 |
brand | +0.421 |
Not | +0.407 |
Imagine | +0.400 |
truly | +0.390 |
Neg logit contributions | |
ried | -0.611 |
ck | -0.581 |
ases | -0.553 |
ies | -0.528 |
aunts | -0.527 |
Token | game |
Token ID | 983 |
Feature activation | -7.40e+00 |
Pos logit contributions | |
brand | +0.101 |
� | +0.092 |
truly | +0.090 |
Not | +0.088 |
Imagine | +0.087 |
Neg logit contributions | |
ried | -0.120 |
ck | -0.112 |
ases | -0.109 |
aunts | -0.102 |
race | -0.098 |
Token | on |
Token ID | 319 |
Feature activation | +4.24e+00 |
Pos logit contributions | |
brand | +0.262 |
truly | +0.243 |
sky | +0.239 |
� | +0.238 |
breathtaking | +0.232 |
Neg logit contributions | |
ried | -0.206 |
ck | -0.191 |
ases | -0.182 |
aunts | -0.170 |
beh | -0.162 |
Token | the |
Token ID | 262 |
Feature activation | -9.37e-01 |
Pos logit contributions | |
ried | +0.097 |
ck | +0.087 |
ases | +0.083 |
aunts | +0.074 |
race | +0.070 |
Neg logit contributions | |
brand | -0.171 |
� | -0.159 |
truly | -0.159 |
sky | -0.155 |
Not | -0.154 |
Token | screen |
Token ID | 3159 |
Feature activation | -7.06e+00 |
Pos logit contributions | |
brand | +0.023 |
� | +0.021 |
truly | +0.020 |
Imagine | +0.020 |
Not | +0.020 |
Neg logit contributions | |
ried | -0.035 |
ck | -0.033 |
ases | -0.032 |
aunts | -0.030 |
race | -0.030 |
Token | girl |
Token ID | 2576 |
Feature activation | -3.76e-01 |
Pos logit contributions | |
brand | +0.020 |
sky | +0.019 |
� | +0.018 |
truly | +0.018 |
Imagine | +0.018 |
Neg logit contributions | |
ried | -0.023 |
ck | -0.021 |
ases | -0.021 |
beh | -0.019 |
aunts | -0.019 |
Token | . |
Token ID | 13 |
Feature activation | +2.32e+00 |
Pos logit contributions | |
brand | +0.010 |
sky | +0.010 |
much | +0.010 |
truly | +0.010 |
Imagine | +0.010 |
Neg logit contributions | |
ried | -0.014 |
ases | -0.012 |
ucks | -0.011 |
ped | -0.011 |
ck | -0.011 |
Token | She |
Token ID | 1375 |
Feature activation | +7.24e-01 |
Pos logit contributions | |
ried | +0.070 |
ases | +0.064 |
ck | +0.061 |
aunts | +0.059 |
race | +0.058 |
Neg logit contributions | |
brand | -0.077 |
truly | -0.069 |
� | -0.066 |
breathtaking | -0.066 |
Imagine | -0.064 |
Token | had |
Token ID | 550 |
Feature activation | -3.24e-01 |
Pos logit contributions | |
ried | +0.026 |
ck | +0.024 |
ases | +0.023 |
aunts | +0.022 |
ped | +0.021 |
Neg logit contributions | |
brand | -0.020 |
� | -0.018 |
truly | -0.018 |
breathtaking | -0.017 |
sky | -0.017 |
Token | a |
Token ID | 257 |
Feature activation | -4.07e+00 |
Pos logit contributions | |
brand | +0.010 |
� | +0.010 |
truly | +0.009 |
breathtaking | +0.009 |
Imagine | +0.009 |
Neg logit contributions | |
ried | -0.010 |
ck | -0.009 |
ases | -0.009 |
aunts | -0.008 |
ped | -0.008 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.72e+01 |
Pos logit contributions | |
brand | +0.120 |
� | +0.114 |
truly | +0.112 |
breathtaking | +0.112 |
Not | +0.111 |
Neg logit contributions | |
ried | -0.116 |
ck | -0.111 |
ases | -0.104 |
ies | -0.102 |
aunts | -0.100 |
Token | new |
Token ID | 649 |
Feature activation | -3.40e+00 |
Pos logit contributions | |
brand | +0.409 |
� | +0.385 |
truly | +0.375 |
Imagine | +0.367 |
Not | +0.366 |
Neg logit contributions | |
ried | -0.647 |
ck | -0.609 |
ases | -0.595 |
aunts | -0.563 |
ies | -0.548 |
Token | plastic |
Token ID | 7309 |
Feature activation | -6.08e+00 |
Pos logit contributions | |
brand | +0.119 |
� | +0.109 |
truly | +0.108 |
sky | +0.106 |
Imagine | +0.104 |
Neg logit contributions | |
ried | -0.097 |
ck | -0.089 |
ases | -0.086 |
aunts | -0.079 |
beh | -0.076 |
Token | bucket |
Token ID | 19236 |
Feature activation | -6.35e+00 |
Pos logit contributions | |
brand | +0.179 |
sky | +0.163 |
� | +0.162 |
truly | +0.154 |
Not | +0.146 |
Neg logit contributions | |
ried | -0.214 |
ck | -0.205 |
ases | -0.190 |
beh | -0.183 |
aunts | -0.180 |
Token | . |
Token ID | 13 |
Feature activation | +2.28e+00 |
Pos logit contributions | |
brand | +0.231 |
� | +0.213 |
truly | +0.212 |
sky | +0.207 |
Imagine | +0.205 |
Neg logit contributions | |
ried | -0.169 |
ck | -0.154 |
ases | -0.149 |
aunts | -0.135 |
race | -0.129 |
Token | She |
Token ID | 1375 |
Feature activation | +6.76e-01 |
Pos logit contributions | |
ases | +0.059 |
ried | +0.059 |
share | +0.055 |
race | +0.054 |
ucks | +0.054 |
Neg logit contributions | |
brand | -0.078 |
truly | -0.070 |
breathtaking | -0.068 |
fantastic | -0.065 |
impossible | -0.064 |
Token | wreck |
Token ID | 15228 |
Feature activation | -6.33e+00 |
Pos logit contributions | |
brand | +0.107 |
� | +0.107 |
Not | +0.103 |
truly | +0.099 |
Cookies | +0.098 |
Neg logit contributions | |
ried | -0.148 |
ck | -0.142 |
steer | -0.135 |
leave | -0.132 |
ran | -0.129 |
Token | , |
Token ID | 11 |
Feature activation | -2.75e+00 |
Pos logit contributions | |
brand | +0.166 |
truly | +0.151 |
sky | +0.146 |
� | +0.143 |
Imagine | +0.140 |
Neg logit contributions | |
ried | -0.232 |
ck | -0.217 |
ases | -0.215 |
aunts | -0.204 |
ucks | -0.200 |
Token | he |
Token ID | 339 |
Feature activation | -8.92e-01 |
Pos logit contributions | |
brand | +0.092 |
� | +0.085 |
truly | +0.083 |
Imagine | +0.081 |
sky | +0.081 |
Neg logit contributions | |
ried | -0.081 |
ck | -0.075 |
ases | -0.072 |
aunts | -0.066 |
race | -0.064 |
Token | saw |
Token ID | 2497 |
Feature activation | +3.22e+00 |
Pos logit contributions | |
brand | +0.024 |
truly | +0.022 |
� | +0.021 |
sky | +0.021 |
Imagine | +0.021 |
Neg logit contributions | |
ried | -0.032 |
ases | -0.030 |
ck | -0.030 |
aunts | -0.028 |
ucks | -0.028 |
Token | a |
Token ID | 257 |
Feature activation | -4.47e+00 |
Pos logit contributions | |
ried | +0.110 |
ck | +0.104 |
ases | +0.101 |
aunts | +0.093 |
ped | +0.091 |
Neg logit contributions | |
brand | -0.090 |
� | -0.088 |
truly | -0.082 |
Imagine | -0.080 |
Not | -0.080 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.72e+01 |
Pos logit contributions | |
� | +0.127 |
Not | +0.124 |
brand | +0.120 |
Eventually | +0.119 |
Finally | +0.119 |
Neg logit contributions | |
ried | -0.124 |
ck | -0.118 |
steer | -0.113 |
leave | -0.110 |
ies | -0.108 |
Token | new |
Token ID | 649 |
Feature activation | -3.51e+00 |
Pos logit contributions | |
� | +0.433 |
brand | +0.417 |
Not | +0.407 |
Imagine | +0.399 |
truly | +0.392 |
Neg logit contributions | |
ried | -0.617 |
ck | -0.585 |
ases | -0.556 |
aunts | -0.531 |
ies | -0.524 |
Token | toy |
Token ID | 13373 |
Feature activation | -4.57e+00 |
Pos logit contributions | |
brand | +0.111 |
� | +0.102 |
truly | +0.101 |
sky | +0.098 |
Not | +0.097 |
Neg logit contributions | |
ried | -0.110 |
ck | -0.102 |
ases | -0.099 |
aunts | -0.091 |
race | -0.088 |
Token | airplane |
Token ID | 19401 |
Feature activation | -2.66e+00 |
Pos logit contributions | |
brand | +0.148 |
sky | +0.136 |
truly | +0.134 |
� | +0.133 |
fact | +0.123 |
Neg logit contributions | |
ried | -0.149 |
ck | -0.140 |
ases | -0.130 |
beh | -0.128 |
aunts | -0.127 |
Token | ! |
Token ID | 0 |
Feature activation | +7.81e-01 |
Pos logit contributions | |
brand | +0.091 |
sky | +0.085 |
truly | +0.084 |
� | +0.081 |
breathtaking | +0.080 |
Neg logit contributions | |
ried | -0.079 |
ases | -0.073 |
ck | -0.070 |
aunts | -0.068 |
ucks | -0.067 |
Token |
|
Token ID | 198 |
Feature activation | -4.72e+00 |
Pos logit contributions | |
ried | +0.021 |
ases | +0.020 |
aunts | +0.018 |
race | +0.018 |
ck | +0.018 |
Neg logit contributions | |
brand | -0.027 |
truly | -0.025 |
breathtaking | -0.024 |
Imagine | -0.023 |
fantastic | -0.023 |
Token | bag |
Token ID | 6131 |
Feature activation | -5.72e+00 |
Pos logit contributions | |
ried | +0.099 |
ck | +0.093 |
ases | +0.090 |
aunts | +0.083 |
beh | +0.080 |
Neg logit contributions | |
brand | -0.104 |
� | -0.093 |
truly | -0.092 |
sky | -0.091 |
Not | -0.088 |
Token | and |
Token ID | 290 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
brand | +0.197 |
truly | +0.181 |
� | +0.181 |
Not | +0.175 |
breathtaking | +0.175 |
Neg logit contributions | |
ried | -0.163 |
ck | -0.148 |
ases | -0.143 |
aunts | -0.132 |
race | -0.127 |
Token | pulled |
Token ID | 5954 |
Feature activation | +6.30e+00 |
Pos logit contributions | |
brand | +0.056 |
truly | +0.051 |
� | +0.049 |
Imagine | +0.047 |
sky | +0.047 |
Neg logit contributions | |
ases | -0.052 |
ried | -0.048 |
ck | -0.045 |
aunts | -0.044 |
ucks | -0.044 |
Token | out |
Token ID | 503 |
Feature activation | +1.10e+00 |
Pos logit contributions | |
ried | +0.207 |
ases | +0.199 |
beh | +0.196 |
aunts | +0.194 |
ck | +0.190 |
Neg logit contributions | |
brand | -0.179 |
sky | -0.166 |
much | -0.165 |
so | -0.160 |
truly | -0.160 |
Token | a |
Token ID | 257 |
Feature activation | -4.51e+00 |
Pos logit contributions | |
ried | +0.041 |
ases | +0.039 |
ck | +0.039 |
aunts | +0.039 |
beh | +0.037 |
Neg logit contributions | |
brand | -0.030 |
much | -0.026 |
sky | -0.025 |
truly | -0.025 |
lot | -0.025 |
Token | brand |
Token ID | 4508 |
Feature activation | -1.72e+01 |
Pos logit contributions | |
brand | +0.125 |
� | +0.122 |
Not | +0.119 |
truly | +0.119 |
Imagine | +0.115 |
Neg logit contributions | |
ried | -0.144 |
ck | -0.136 |
ases | -0.125 |
steer | -0.119 |
aunts | -0.118 |
Token | new |
Token ID | 649 |
Feature activation | -3.53e+00 |
Pos logit contributions | |
brand | +0.448 |
� | +0.428 |
truly | +0.413 |
Imagine | +0.407 |
Not | +0.407 |
Neg logit contributions | |
ried | -0.608 |
ck | -0.578 |
ases | -0.558 |
aunts | -0.525 |
ucks | -0.514 |
Token | doll |
Token ID | 3654 |
Feature activation | -3.37e-01 |
Pos logit contributions | |
brand | +0.110 |
� | +0.095 |
truly | +0.094 |
sky | +0.093 |
breathtaking | +0.092 |
Neg logit contributions | |
ried | -0.120 |
ck | -0.111 |
ases | -0.110 |
aunts | -0.101 |
beh | -0.100 |
Token | for |
Token ID | 329 |
Feature activation | +4.27e+00 |
Pos logit contributions | |
brand | +0.012 |
� | +0.011 |
truly | +0.011 |
sky | +0.011 |
case | +0.010 |
Neg logit contributions | |
ried | -0.009 |
ases | -0.009 |
ck | -0.008 |
aunts | -0.008 |
ucks | -0.008 |
Token | Sally |
Token ID | 25737 |
Feature activation | +8.11e+00 |
Pos logit contributions | |
ried | +0.118 |
ases | +0.102 |
aunts | +0.098 |
ucks | +0.097 |
ck | +0.096 |
Neg logit contributions | |
brand | -0.163 |
� | -0.151 |
truly | -0.149 |
Not | -0.144 |
sky | -0.143 |
Token | . |
Token ID | 13 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
ases | +0.235 |
ried | +0.231 |
ck | +0.223 |
ucks | +0.220 |
aunts | +0.210 |
Neg logit contributions | |
brand | -0.276 |
truly | -0.260 |
sky | -0.244 |
Not | -0.243 |
� | -0.240 |
Token | their |
Token ID | 511 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
ried | +0.188 |
beh | +0.171 |
ck | +0.156 |
ases | +0.155 |
ies | +0.150 |
Neg logit contributions | |
brand | -0.191 |
� | -0.175 |
lot | -0.174 |
truly | -0.172 |
Not | -0.171 |
Token | yard |
Token ID | 12699 |
Feature activation | -5.13e+00 |
Pos logit contributions | |
ried | +0.069 |
ck | +0.067 |
beh | +0.064 |
ases | +0.063 |
ped | +0.061 |
Neg logit contributions | |
brand | -0.051 |
sky | -0.047 |
lot | -0.042 |
fact | -0.039 |
most | -0.039 |
Token | , |
Token ID | 11 |
Feature activation | -2.14e+00 |
Pos logit contributions | |
brand | +0.158 |
� | +0.146 |
sky | +0.146 |
truly | +0.139 |
Not | +0.135 |
Neg logit contributions | |
ried | -0.161 |
ck | -0.154 |
ases | -0.149 |
beh | -0.139 |
aunts | -0.138 |
Token | crying |
Token ID | 13774 |
Feature activation | +3.07e+00 |
Pos logit contributions | |
brand | +0.066 |
truly | +0.060 |
� | +0.059 |
sky | +0.058 |
Not | +0.058 |
Neg logit contributions | |
ried | -0.069 |
ck | -0.064 |
ases | -0.063 |
aunts | -0.058 |
beh | -0.056 |
Token | and |
Token ID | 290 |
Feature activation | -8.52e-01 |
Pos logit contributions | |
ried | +0.086 |
ck | +0.083 |
ases | +0.078 |
race | +0.076 |
beh | +0.074 |
Neg logit contributions | |
brand | -0.104 |
sky | -0.096 |
truly | -0.095 |
� | -0.091 |
Not | -0.091 |
Token | holding |
Token ID | 4769 |
Feature activation | +8.73e+00 |
Pos logit contributions | |
brand | +0.026 |
sky | +0.023 |
truly | +0.023 |
fact | +0.022 |
Imagine | +0.021 |
Neg logit contributions | |
ried | -0.026 |
ases | -0.025 |
ck | -0.025 |
ucks | -0.024 |
aunts | -0.023 |
Token | each |
Token ID | 1123 |
Feature activation | +1.96e+00 |
Pos logit contributions | |
ried | +0.240 |
ck | +0.210 |
race | +0.209 |
aunts | +0.208 |
beh | +0.201 |
Neg logit contributions | |
brand | -0.284 |
lot | -0.281 |
most | -0.276 |
fantastic | -0.260 |
sky | -0.260 |
Token | other |
Token ID | 584 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
ried | +0.082 |
ck | +0.072 |
beh | +0.072 |
ases | +0.069 |
rying | +0.069 |
Neg logit contributions | |
sky | -0.041 |
lot | -0.041 |
brand | -0.039 |
Not | -0.037 |
fact | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | +2.03e+00 |
Pos logit contributions | |
ried | +0.004 |
ases | +0.004 |
aunts | +0.004 |
ck | +0.004 |
ucks | +0.004 |
Neg logit contributions | |
brand | -0.003 |
sky | -0.003 |
much | -0.003 |
truly | -0.003 |
Not | -0.002 |
Token | They |
Token ID | 1119 |
Feature activation | +5.07e-01 |
Pos logit contributions | |
ried | +0.062 |
ases | +0.057 |
ck | +0.056 |
aunts | +0.053 |
pered | +0.050 |
Neg logit contributions | |
brand | -0.065 |
truly | -0.059 |
breathtaking | -0.057 |
� | -0.056 |
Imagine | -0.054 |
Token | did |
Token ID | 750 |
Feature activation | -4.09e+00 |
Pos logit contributions | |
ried | +0.019 |
ases | +0.018 |
ucks | +0.017 |
ck | +0.017 |
race | +0.017 |
Neg logit contributions | |
brand | -0.013 |
truly | -0.012 |
sky | -0.011 |
� | -0.011 |
Imagine | -0.011 |
Token |
|
Token ID | 198 |
Feature activation | -4.09e+00 |
Pos logit contributions | |
brand | +0.236 |
� | +0.205 |
truly | +0.201 |
Not | +0.193 |
sky | +0.191 |
Neg logit contributions | |
ried | -0.519 |
ck | -0.492 |
ases | -0.481 |
aunts | -0.452 |
beh | -0.446 |
Token |
|
Token ID | 198 |
Feature activation | -4.17e+00 |
Pos logit contributions | |
brand | +0.113 |
� | +0.103 |
truly | +0.101 |
breathtaking | +0.101 |
fact | +0.099 |
Neg logit contributions | |
ried | -0.137 |
ck | -0.132 |
aunts | -0.125 |
ases | -0.125 |
ucks | -0.120 |
Token | After |
Token ID | 3260 |
Feature activation | -3.04e+00 |
Pos logit contributions | |
brand | +0.101 |
� | +0.096 |
truly | +0.089 |
sky | +0.087 |
fact | +0.085 |
Neg logit contributions | |
ried | -0.157 |
ck | -0.153 |
aunts | -0.143 |
ases | -0.141 |
steer | -0.138 |
Token | dinner |
Token ID | 8073 |
Feature activation | -4.99e+00 |
Pos logit contributions | |
brand | +0.103 |
truly | +0.096 |
Imagine | +0.094 |
breathtaking | +0.094 |
� | +0.094 |
Neg logit contributions | |
ried | -0.083 |
ck | -0.079 |
ases | -0.075 |
aunts | -0.068 |
race | -0.067 |
Token | , |
Token ID | 11 |
Feature activation | -2.07e+00 |
Pos logit contributions | |
brand | +0.186 |
truly | +0.171 |
breathtaking | +0.170 |
Imagine | +0.162 |
daunting | +0.161 |
Neg logit contributions | |
ried | -0.118 |
ck | -0.109 |
ases | -0.108 |
ie | -0.101 |
leave | -0.098 |
Token | Tim |
Token ID | 5045 |
Feature activation | +9.67e+00 |
Pos logit contributions | |
brand | +0.097 |
truly | +0.091 |
� | +0.091 |
sky | +0.089 |
breathtaking | +0.089 |
Neg logit contributions | |
ried | -0.033 |
ck | -0.028 |
ases | -0.026 |
aunts | -0.022 |
race | -0.019 |
Token | my |
Token ID | 1820 |
Feature activation | +4.21e+00 |
Pos logit contributions | |
ried | +0.291 |
ck | +0.280 |
ases | +0.276 |
beh | +0.246 |
pered | +0.241 |
Neg logit contributions | |
brand | -0.305 |
sky | -0.290 |
Imagine | -0.288 |
truly | -0.287 |
� | -0.268 |
Token | and |
Token ID | 290 |
Feature activation | -7.11e-01 |
Pos logit contributions | |
ried | +0.138 |
ases | +0.133 |
ck | +0.127 |
beh | +0.120 |
aunts | +0.117 |
Neg logit contributions | |
brand | -0.126 |
truly | -0.118 |
sky | -0.118 |
Imagine | -0.113 |
� | -0.112 |
Token | his |
Token ID | 465 |
Feature activation | +2.78e+00 |
Pos logit contributions | |
brand | +0.023 |
Not | +0.022 |
sky | +0.021 |
lot | +0.021 |
so | +0.021 |
Neg logit contributions | |
ried | -0.022 |
ases | -0.020 |
ck | -0.019 |
pered | -0.018 |
rying | -0.018 |
Token | family |
Token ID | 1641 |
Feature activation | +6.30e-01 |
Pos logit contributions | |
ried | +0.101 |
ck | +0.092 |
ases | +0.091 |
race | +0.088 |
beh | +0.086 |
Neg logit contributions | |
brand | -0.083 |
sky | -0.079 |
truly | -0.074 |
most | -0.069 |
lot | -0.069 |
Token | went |
Token ID | 1816 |
Feature activation | +8.15e+00 |
Pos logit contributions | |
ases | +0.021 |
ck | +0.019 |
ried | +0.019 |
ucks | +0.019 |
beh | +0.018 |
Neg logit contributions | |
truly | -0.019 |
sky | -0.018 |
Imagine | -0.018 |
brand | -0.018 |
� | -0.016 |
Token | there |
Token ID | 612 |
Feature activation | -2.97e+00 |
Pos logit contributions | |
brand | +0.118 |
� | +0.107 |
truly | +0.105 |
breathtaking | +0.102 |
Imagine | +0.101 |
Neg logit contributions | |
ried | -0.137 |
ck | -0.127 |
ases | -0.124 |
aunts | -0.116 |
ies | -0.113 |
Token | was |
Token ID | 373 |
Feature activation | -3.54e+00 |
Pos logit contributions | |
brand | +0.096 |
� | +0.084 |
Finally | +0.083 |
daunting | +0.083 |
breathtaking | +0.082 |
Neg logit contributions | |
ried | -0.087 |
ie | -0.079 |
ck | -0.079 |
him | -0.078 |
his | -0.077 |
Token | a |
Token ID | 257 |
Feature activation | -3.68e+00 |
Pos logit contributions | |
brand | +0.119 |
� | +0.111 |
truly | +0.107 |
Imagine | +0.105 |
daunting | +0.104 |
Neg logit contributions | |
ried | -0.095 |
ck | -0.094 |
ases | -0.090 |
ped | -0.080 |
aunts | -0.080 |
Token | girl |
Token ID | 2576 |
Feature activation | +1.65e-01 |
Pos logit contributions | |
brand | +0.095 |
� | +0.084 |
truly | +0.084 |
sky | +0.082 |
Imagine | +0.080 |
Neg logit contributions | |
ried | -0.144 |
ases | -0.133 |
ck | -0.131 |
beh | -0.120 |
ped | -0.120 |
Token | called |
Token ID | 1444 |
Feature activation | +7.75e+00 |
Pos logit contributions | |
ried | +0.005 |
ases | +0.004 |
ck | +0.004 |
ped | +0.004 |
ucks | +0.004 |
Neg logit contributions | |
brand | -0.005 |
sky | -0.005 |
Imagine | -0.005 |
truly | -0.005 |
much | -0.005 |
Token | Sally |
Token ID | 25737 |
Feature activation | +9.04e+00 |
Pos logit contributions | |
ried | +0.230 |
ck | +0.214 |
ases | +0.202 |
aunts | +0.193 |
ucks | +0.187 |
Neg logit contributions | |
� | -0.252 |
brand | -0.252 |
Not | -0.248 |
sky | -0.243 |
Cookies | -0.235 |
Token | who |
Token ID | 508 |
Feature activation | -6.83e-01 |
Pos logit contributions | |
ried | +0.237 |
ases | +0.220 |
ck | +0.209 |
beh | +0.188 |
aunts | +0.183 |
Neg logit contributions | |
brand | -0.330 |
sky | -0.318 |
truly | -0.317 |
Imagine | -0.299 |
Not | -0.299 |
Token | liked |
Token ID | 8288 |
Feature activation | -4.66e-01 |
Pos logit contributions | |
brand | +0.022 |
� | +0.021 |
breathtaking | +0.019 |
Finally | +0.019 |
truly | +0.019 |
Neg logit contributions | |
ried | -0.019 |
ck | -0.019 |
aunts | -0.017 |
ases | -0.017 |
share | -0.017 |
Token | to |
Token ID | 284 |
Feature activation | +6.17e+00 |
Pos logit contributions | |
brand | +0.018 |
� | +0.017 |
truly | +0.017 |
sky | +0.017 |
Imagine | +0.017 |
Neg logit contributions | |
ried | -0.011 |
ck | -0.010 |
ases | -0.009 |
aunts | -0.008 |
race | -0.008 |
Token | play |
Token ID | 711 |
Feature activation | +3.57e+00 |
Pos logit contributions | |
ried | +0.186 |
ck | +0.155 |
ases | +0.148 |
aunts | +0.143 |
ucks | +0.141 |
Neg logit contributions | |
brand | -0.220 |
� | -0.209 |
truly | -0.203 |
Not | -0.203 |
sky | -0.202 |
Token | in |
Token ID | 287 |
Feature activation | +6.02e+00 |
Pos logit contributions | |
ried | +0.116 |
ck | +0.099 |
ases | +0.091 |
beh | +0.090 |
aunts | +0.087 |
Neg logit contributions | |
brand | -0.124 |
� | -0.117 |
sky | -0.114 |
truly | -0.110 |
Imagine | -0.109 |
Token | and |
Token ID | 290 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
ried | +0.144 |
ases | +0.140 |
ck | +0.137 |
ucks | +0.134 |
beh | +0.134 |
Neg logit contributions | |
brand | -0.099 |
� | -0.089 |
sky | -0.087 |
so | -0.082 |
truly | -0.082 |
Token | tick |
Token ID | 4378 |
Feature activation | +3.01e+00 |
Pos logit contributions | |
brand | +0.038 |
so | +0.033 |
Not | +0.033 |
sky | +0.033 |
fact | +0.033 |
Neg logit contributions | |
ried | -0.029 |
ases | -0.028 |
ucks | -0.027 |
beh | -0.026 |
rying | -0.025 |
Token | led |
Token ID | 992 |
Feature activation | +2.99e+00 |
Pos logit contributions | |
ried | +0.076 |
race | +0.070 |
aunts | +0.069 |
adopt | +0.066 |
kidnap | +0.066 |
Neg logit contributions | |
brand | -0.090 |
� | -0.085 |
fact | -0.081 |
Not | -0.081 |
Too | -0.081 |
Token | them |
Token ID | 606 |
Feature activation | -3.18e-01 |
Pos logit contributions | |
ried | +0.071 |
ucks | +0.064 |
aunts | +0.060 |
ck | +0.059 |
ases | +0.058 |
Neg logit contributions | |
brand | -0.109 |
so | -0.105 |
much | -0.100 |
sky | -0.098 |
Not | -0.097 |
Token | . |
Token ID | 13 |
Feature activation | +1.92e+00 |
Pos logit contributions | |
brand | +0.012 |
truly | +0.011 |
� | +0.010 |
sky | +0.010 |
Not | +0.010 |
Neg logit contributions | |
ried | -0.008 |
ases | -0.008 |
ck | -0.008 |
ucks | -0.008 |
aunts | -0.007 |
Token | Tom |
Token ID | 4186 |
Feature activation | +9.78e+00 |
Pos logit contributions | |
ried | +0.042 |
ases | +0.037 |
ck | +0.036 |
aunts | +0.033 |
race | +0.031 |
Neg logit contributions | |
brand | -0.078 |
truly | -0.073 |
breathtaking | -0.072 |
� | -0.071 |
Imagine | -0.070 |
Token | laughed |
Token ID | 13818 |
Feature activation | +2.26e+00 |
Pos logit contributions | |
ried | +0.318 |
ases | +0.295 |
ck | +0.291 |
aunts | +0.268 |
ucks | +0.267 |
Neg logit contributions | |
brand | -0.292 |
truly | -0.270 |
sky | -0.262 |
� | -0.262 |
Imagine | -0.261 |
Token | and |
Token ID | 290 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
ried | +0.069 |
ck | +0.064 |
ases | +0.062 |
aunts | +0.057 |
beh | +0.055 |
Neg logit contributions | |
brand | -0.074 |
truly | -0.067 |
� | -0.067 |
sky | -0.066 |
Not | -0.065 |
Token | squ |
Token ID | 2809 |
Feature activation | +3.12e+00 |
Pos logit contributions | |
brand | +0.037 |
truly | +0.034 |
� | +0.033 |
sky | +0.033 |
Imagine | +0.032 |
Neg logit contributions | |
ried | -0.027 |
ases | -0.027 |
ucks | -0.024 |
ck | -0.024 |
aunts | -0.023 |
Token | irmed |
Token ID | 15491 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
ried | +0.084 |
steer | +0.075 |
aunts | +0.074 |
pa | +0.074 |
race | +0.072 |
Neg logit contributions | |
brand | -0.095 |
� | -0.086 |
Imagine | -0.085 |
Eventually | -0.085 |
Finally | -0.083 |
Token | . |
Token ID | 13 |
Feature activation | +2.02e+00 |
Pos logit contributions | |
ried | +0.075 |
ck | +0.071 |
ases | +0.069 |
aunts | +0.065 |
beh | +0.063 |
Neg logit contributions | |
brand | -0.076 |
truly | -0.069 |
sky | -0.068 |
� | -0.068 |
Imagine | -0.066 |
Token | them |
Token ID | 606 |
Feature activation | -2.00e-01 |
Pos logit contributions | |
ried | +0.333 |
beh | +0.325 |
rying | +0.321 |
maneu | +0.299 |
ci | +0.299 |
Neg logit contributions | |
lot | -0.264 |
much | -0.244 |
sky | -0.240 |
so | -0.225 |
Not | -0.216 |
Token | . |
Token ID | 13 |
Feature activation | +2.23e+00 |
Pos logit contributions | |
brand | +0.006 |
much | +0.006 |
sky | +0.006 |
so | +0.006 |
Not | +0.006 |
Neg logit contributions | |
ried | -0.005 |
ucks | -0.005 |
pa | -0.005 |
ck | -0.005 |
ases | -0.005 |
Token |
|
Token ID | 198 |
Feature activation | -4.06e+00 |
Pos logit contributions | |
ried | +0.075 |
ck | +0.070 |
ases | +0.069 |
aunts | +0.064 |
ucks | +0.061 |
Neg logit contributions | |
brand | -0.066 |
truly | -0.059 |
� | -0.059 |
breathtaking | -0.057 |
Imagine | -0.056 |
Token |
|
Token ID | 198 |
Feature activation | -4.13e+00 |
Pos logit contributions | |
brand | +0.080 |
� | +0.069 |
truly | +0.069 |
sky | +0.066 |
Not | +0.065 |
Neg logit contributions | |
ried | -0.176 |
ck | -0.165 |
ases | -0.162 |
aunts | -0.153 |
beh | -0.151 |
Token | " |
Token ID | 1 |
Feature activation | -2.10e-01 |
Pos logit contributions | |
brand | +0.119 |
Finally | +0.113 |
truly | +0.112 |
Imagine | +0.111 |
Not | +0.111 |
Neg logit contributions | |
ried | -0.136 |
ases | -0.122 |
ck | -0.122 |
aunts | -0.111 |
pa | -0.111 |
Token | Help |
Token ID | 22087 |
Feature activation | +7.83e+00 |
Pos logit contributions | |
brand | +0.006 |
Imagine | +0.006 |
� | +0.006 |
sky | +0.006 |
truly | +0.006 |
Neg logit contributions | |
ried | -0.007 |
ck | -0.006 |
ases | -0.006 |
ies | -0.006 |
aunts | -0.006 |
Token | , |
Token ID | 11 |
Feature activation | -2.54e+00 |
Pos logit contributions | |
ried | +0.338 |
beh | +0.319 |
ck | +0.301 |
ases | +0.292 |
aunts | +0.292 |
Neg logit contributions | |
sky | -0.137 |
fact | -0.131 |
so | -0.122 |
brand | -0.118 |
� | -0.115 |
Token | help |
Token ID | 1037 |
Feature activation | +4.76e+00 |
Pos logit contributions | |
brand | +0.079 |
Not | +0.078 |
truly | +0.074 |
lot | +0.074 |
fact | +0.072 |
Neg logit contributions | |
ried | -0.077 |
ck | -0.068 |
ases | -0.068 |
aunts | -0.067 |
beh | -0.067 |
Token | , |
Token ID | 11 |
Feature activation | -2.50e+00 |
Pos logit contributions | |
ried | +0.157 |
ck | +0.145 |
ases | +0.142 |
beh | +0.138 |
race | +0.134 |
Neg logit contributions | |
brand | -0.131 |
truly | -0.122 |
sky | -0.122 |
Not | -0.119 |
Imagine | -0.119 |
Token | a |
Token ID | 257 |
Feature activation | -4.18e+00 |
Pos logit contributions | |
brand | +0.077 |
Not | +0.074 |
truly | +0.071 |
sky | +0.070 |
lot | +0.069 |
Neg logit contributions | |
ried | -0.079 |
ases | -0.071 |
ck | -0.070 |
aunts | -0.068 |
beh | -0.067 |
Token | fierce |
Token ID | 14800 |
Feature activation | +3.35e-01 |
Pos logit contributions | |
brand | +0.130 |
� | +0.121 |
truly | +0.119 |
breathtaking | +0.114 |
Not | +0.114 |
Neg logit contributions | |
ried | -0.130 |
ck | -0.122 |
ases | -0.117 |
aunts | -0.109 |
race | -0.104 |
Token | are |
Token ID | 389 |
Feature activation | -5.94e+00 |
Pos logit contributions | |
ried | +0.408 |
ck | +0.373 |
ucks | +0.363 |
beh | +0.362 |
aunts | +0.359 |
Neg logit contributions | |
truly | -0.225 |
sky | -0.223 |
Imagine | -0.221 |
may | -0.220 |
brand | -0.218 |
Token | brothers |
Token ID | 9397 |
Feature activation | +7.47e+00 |
Pos logit contributions | |
brand | +0.217 |
truly | +0.200 |
� | +0.197 |
sky | +0.195 |
Not | +0.194 |
Neg logit contributions | |
ried | -0.169 |
ck | -0.152 |
ases | -0.138 |
aunts | -0.131 |
steer | -0.128 |
Token | . |
Token ID | 13 |
Feature activation | +3.21e+00 |
Pos logit contributions | |
ried | +0.233 |
ck | +0.201 |
ases | +0.199 |
beh | +0.193 |
ped | +0.187 |
Neg logit contributions | |
brand | -0.254 |
sky | -0.247 |
truly | -0.246 |
� | -0.244 |
Not | -0.230 |
Token | They |
Token ID | 1119 |
Feature activation | +1.00e+00 |
Pos logit contributions | |
ried | +0.110 |
ck | +0.104 |
ases | +0.098 |
aunts | +0.092 |
beh | +0.091 |
Neg logit contributions | |
brand | -0.092 |
� | -0.087 |
sky | -0.084 |
truly | -0.084 |
Not | -0.083 |
Token | like |
Token ID | 588 |
Feature activation | -8.37e-01 |
Pos logit contributions | |
ried | +0.042 |
ck | +0.039 |
ases | +0.038 |
ucks | +0.037 |
aunts | +0.037 |
Neg logit contributions | |
truly | -0.021 |
brand | -0.020 |
Not | -0.020 |
sky | -0.018 |
definitely | -0.018 |
Token | to |
Token ID | 284 |
Feature activation | +6.46e+00 |
Pos logit contributions | |
brand | +0.026 |
sky | +0.025 |
Not | +0.025 |
truly | +0.024 |
Imagine | +0.023 |
Neg logit contributions | |
ried | -0.027 |
ck | -0.026 |
ases | -0.024 |
aunts | -0.022 |
ped | -0.022 |
Token | play |
Token ID | 711 |
Feature activation | +3.73e+00 |
Pos logit contributions | |
ried | +0.221 |
beh | +0.179 |
ck | +0.178 |
ucks | +0.177 |
rying | +0.176 |
Neg logit contributions | |
� | -0.206 |
brand | -0.200 |
truly | -0.192 |
sky | -0.192 |
Not | -0.190 |
Token | in |
Token ID | 287 |
Feature activation | +6.11e+00 |
Pos logit contributions | |
ried | +0.113 |
ck | +0.098 |
beh | +0.094 |
ases | +0.090 |
aunts | +0.089 |
Neg logit contributions | |
brand | -0.132 |
� | -0.121 |
truly | -0.119 |
sky | -0.118 |
Imagine | -0.115 |
Token | the |
Token ID | 262 |
Feature activation | +7.66e-02 |
Pos logit contributions | |
ried | +0.145 |
ck | +0.126 |
ases | +0.121 |
aunts | +0.109 |
beh | +0.106 |
Neg logit contributions | |
brand | -0.247 |
� | -0.227 |
truly | -0.227 |
sky | -0.225 |
breathtaking | -0.222 |
Token | yard |
Token ID | 12699 |
Feature activation | -4.12e+00 |
Pos logit contributions | |
ried | +0.003 |
ck | +0.003 |
ases | +0.003 |
aunts | +0.003 |
race | +0.003 |
Neg logit contributions | |
brand | -0.002 |
� | -0.001 |
truly | -0.001 |
Not | -0.001 |
sky | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +3.25e+00 |
Pos logit contributions | |
brand | +0.180 |
� | +0.169 |
truly | +0.169 |
sky | +0.166 |
Not | +0.164 |
Neg logit contributions | |
ried | -0.079 |
ck | -0.070 |
ases | -0.066 |
aunts | -0.057 |
beh | -0.054 |
Token | had |
Token ID | 550 |
Feature activation | -7.15e-01 |
Pos logit contributions | |
brand | +0.005 |
truly | +0.005 |
� | +0.004 |
breathtaking | +0.004 |
sky | +0.004 |
Neg logit contributions | |
ried | -0.006 |
ases | -0.006 |
ck | -0.006 |
ucks | -0.005 |
aunts | -0.005 |
Token | used |
Token ID | 973 |
Feature activation | +5.59e+00 |
Pos logit contributions | |
brand | +0.022 |
� | +0.022 |
Not | +0.020 |
Imagine | +0.020 |
sky | +0.019 |
Neg logit contributions | |
ried | -0.022 |
ck | -0.021 |
ases | -0.019 |
ped | -0.018 |
share | -0.017 |
Token | the |
Token ID | 262 |
Feature activation | -7.05e-01 |
Pos logit contributions | |
ried | +0.120 |
ck | +0.108 |
ases | +0.102 |
aunts | +0.084 |
ped | +0.083 |
Neg logit contributions | |
brand | -0.226 |
� | -0.220 |
Imagine | -0.215 |
truly | -0.211 |
Not | -0.211 |
Token | melted |
Token ID | 24178 |
Feature activation | +2.85e+00 |
Pos logit contributions | |
� | +0.022 |
brand | +0.022 |
Not | +0.021 |
Imagine | +0.021 |
truly | +0.020 |
Neg logit contributions | |
ried | -0.021 |
ck | -0.019 |
ases | -0.018 |
steer | -0.017 |
ies | -0.017 |
Token | ocean |
Token ID | 9151 |
Feature activation | -9.01e+00 |
Pos logit contributions | |
ried | +0.103 |
ck | +0.097 |
ases | +0.094 |
aunts | +0.090 |
beh | +0.088 |
Neg logit contributions | |
brand | -0.077 |
sky | -0.067 |
truly | -0.067 |
� | -0.067 |
breathtaking | -0.065 |
Token | to |
Token ID | 284 |
Feature activation | +5.93e+00 |
Pos logit contributions | |
brand | +0.310 |
truly | +0.283 |
� | +0.282 |
sky | +0.274 |
breathtaking | +0.272 |
Neg logit contributions | |
ried | -0.258 |
ck | -0.238 |
ases | -0.232 |
aunts | -0.214 |
race | -0.204 |
Token | make |
Token ID | 787 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
ried | +0.150 |
ck | +0.132 |
ases | +0.131 |
aunts | +0.121 |
beh | +0.115 |
Neg logit contributions | |
brand | -0.227 |
truly | -0.211 |
� | -0.209 |
breathtaking | -0.204 |
sky | -0.204 |
Token | a |
Token ID | 257 |
Feature activation | -4.17e+00 |
Pos logit contributions | |
ried | +0.050 |
ck | +0.045 |
ases | +0.044 |
aunts | +0.042 |
race | +0.040 |
Neg logit contributions | |
brand | -0.053 |
truly | -0.048 |
breathtaking | -0.046 |
� | -0.046 |
sky | -0.046 |
Token | beautiful |
Token ID | 4950 |
Feature activation | -5.24e+00 |
Pos logit contributions | |
� | +0.104 |
Not | +0.099 |
brand | +0.099 |
Imagine | +0.097 |
Too | +0.091 |
Neg logit contributions | |
ried | -0.142 |
ck | -0.140 |
ases | -0.127 |
steer | -0.126 |
ies | -0.125 |
Token | sand |
Token ID | 6450 |
Feature activation | -7.06e+00 |
Pos logit contributions | |
� | +0.138 |
brand | +0.138 |
Not | +0.135 |
Imagine | +0.130 |
truly | +0.127 |
Neg logit contributions | |
ried | -0.173 |
ck | -0.168 |
ases | -0.156 |
ies | -0.155 |
ie | -0.152 |
Token | castle |
Token ID | 18676 |
Feature activation | -1.81e+00 |
Pos logit contributions | |
brand | +0.200 |
truly | +0.179 |
sky | +0.173 |
� | +0.173 |
breathtaking | +0.169 |
Neg logit contributions | |
ried | -0.250 |
ck | -0.234 |
ases | -0.229 |
aunts | -0.217 |
beh | -0.214 |
Token |
|
Token ID | 198 |
Feature activation | -3.46e+00 |
Pos logit contributions | |
ried | +0.091 |
ases | +0.085 |
ck | +0.082 |
aunts | +0.079 |
ucks | +0.078 |
Neg logit contributions | |
brand | -0.072 |
truly | -0.064 |
breathtaking | -0.063 |
� | -0.061 |
fact | -0.058 |
Token |
|
Token ID | 198 |
Feature activation | -3.52e+00 |
Pos logit contributions | |
brand | +0.071 |
� | +0.061 |
truly | +0.061 |
sky | +0.058 |
Not | +0.057 |
Neg logit contributions | |
ried | -0.148 |
ck | -0.139 |
ases | -0.136 |
aunts | -0.128 |
beh | -0.126 |
Token | She |
Token ID | 3347 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
brand | +0.101 |
Finally | +0.094 |
� | +0.093 |
Imagine | +0.092 |
Not | +0.091 |
Neg logit contributions | |
ried | -0.122 |
ck | -0.109 |
ases | -0.108 |
ped | -0.099 |
aunts | -0.098 |
Token | sighed |
Token ID | 21893 |
Feature activation | +5.00e-01 |
Pos logit contributions | |
ried | +0.037 |
ases | +0.036 |
ck | +0.033 |
beh | +0.033 |
aunts | +0.032 |
Neg logit contributions | |
brand | -0.045 |
truly | -0.041 |
sky | -0.040 |
Imagine | -0.040 |
� | -0.038 |
Token | . |
Token ID | 13 |
Feature activation | +2.72e+00 |
Pos logit contributions | |
ried | +0.012 |
ases | +0.011 |
ck | +0.011 |
aunts | +0.010 |
beh | +0.009 |
Neg logit contributions | |
brand | -0.020 |
truly | -0.018 |
� | -0.018 |
sky | -0.018 |
Not | -0.018 |
Token | She |
Token ID | 1375 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
ried | +0.071 |
ases | +0.066 |
aunts | +0.060 |
pered | +0.059 |
ucks | +0.059 |
Neg logit contributions | |
brand | -0.097 |
truly | -0.091 |
breathtaking | -0.088 |
fantastic | -0.084 |
� | -0.084 |
Token | smiled |
Token ID | 13541 |
Feature activation | +2.70e+00 |
Pos logit contributions | |
ried | +0.054 |
ases | +0.050 |
ck | +0.050 |
aunts | +0.046 |
race | +0.045 |
Neg logit contributions | |
brand | -0.043 |
truly | -0.039 |
� | -0.038 |
sky | -0.037 |
Imagine | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | +2.87e+00 |
Pos logit contributions | |
ried | +0.079 |
ck | +0.075 |
ases | +0.070 |
beh | +0.065 |
race | +0.063 |
Neg logit contributions | |
brand | -0.095 |
sky | -0.088 |
truly | -0.086 |
Not | -0.083 |
Imagine | -0.082 |
Token | She |
Token ID | 1375 |
Feature activation | +1.63e+00 |
Pos logit contributions | |
ried | +0.088 |
ases | +0.083 |
aunts | +0.076 |
ucks | +0.076 |
ck | +0.075 |
Neg logit contributions | |
brand | -0.089 |
truly | -0.082 |
breathtaking | -0.079 |
� | -0.077 |
fantastic | -0.075 |
Token | said |
Token ID | 531 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
ried | +0.053 |
ck | +0.049 |
ases | +0.048 |
aunts | +0.044 |
race | +0.043 |
Neg logit contributions | |
brand | -0.050 |
truly | -0.046 |
� | -0.046 |
sky | -0.044 |
Imagine | -0.044 |
Token | , |
Token ID | 11 |
Feature activation | -1.13e+00 |
Pos logit contributions | |
ried | +0.221 |
ck | +0.207 |
ases | +0.198 |
aunts | +0.191 |
beh | +0.186 |
Neg logit contributions | |
brand | -0.147 |
Not | -0.133 |
sky | -0.132 |
� | -0.131 |
truly | -0.128 |
Token | . |
Token ID | 13 |
Feature activation | +2.44e+00 |
Pos logit contributions | |
brand | +0.112 |
truly | +0.104 |
� | +0.104 |
sky | +0.102 |
Imagine | +0.100 |
Neg logit contributions | |
ried | -0.056 |
ck | -0.050 |
ases | -0.047 |
aunts | -0.043 |
beh | -0.040 |
Token | They |
Token ID | 1119 |
Feature activation | +5.68e-01 |
Pos logit contributions | |
ried | +0.076 |
ck | +0.069 |
ases | +0.069 |
aunts | +0.063 |
ucks | +0.061 |
Neg logit contributions | |
brand | -0.079 |
truly | -0.072 |
� | -0.071 |
breathtaking | -0.069 |
Imagine | -0.069 |
Token | saw |
Token ID | 2497 |
Feature activation | +3.89e+00 |
Pos logit contributions | |
ried | +0.022 |
ases | +0.021 |
ck | +0.021 |
ucks | +0.020 |
beh | +0.020 |
Neg logit contributions | |
brand | -0.013 |
truly | -0.012 |
Imagine | -0.011 |
sky | -0.011 |
definitely | -0.011 |
Token | cars |
Token ID | 5006 |
Feature activation | +6.16e-02 |
Pos logit contributions | |
ried | +0.124 |
ck | +0.116 |
ases | +0.114 |
aunts | +0.104 |
ped | +0.101 |
Neg logit contributions | |
brand | -0.113 |
� | -0.113 |
truly | -0.107 |
Imagine | -0.104 |
Not | -0.103 |
Token | , |
Token ID | 11 |
Feature activation | -1.80e+00 |
Pos logit contributions | |
ried | +0.002 |
ck | +0.002 |
ases | +0.002 |
aunts | +0.002 |
beh | +0.002 |
Neg logit contributions | |
brand | -0.002 |
� | -0.002 |
truly | -0.002 |
sky | -0.002 |
Imagine | -0.002 |
Token | dolls |
Token ID | 36062 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
brand | +0.071 |
� | +0.070 |
truly | +0.067 |
breathtaking | +0.065 |
Imagine | +0.064 |
Neg logit contributions | |
ried | -0.041 |
ck | -0.038 |
ases | -0.036 |
aunts | -0.031 |
steer | -0.031 |
Token | , |
Token ID | 11 |
Feature activation | -1.79e+00 |
Pos logit contributions | |
ried | +0.033 |
ck | +0.031 |
ases | +0.030 |
aunts | +0.027 |
ies | +0.027 |
Neg logit contributions | |
brand | -0.037 |
truly | -0.034 |
� | -0.033 |
breathtaking | -0.033 |
Imagine | -0.032 |
Token | blocks |
Token ID | 7021 |
Feature activation | -3.36e+00 |
Pos logit contributions | |
brand | +0.067 |
� | +0.064 |
truly | +0.063 |
breathtaking | +0.061 |
Imagine | +0.060 |
Neg logit contributions | |
ried | -0.045 |
ck | -0.041 |
ases | -0.040 |
aunts | -0.035 |
ies | -0.034 |
Token | , |
Token ID | 11 |
Feature activation | -1.76e+00 |
Pos logit contributions | |
brand | +0.120 |
truly | +0.110 |
� | +0.109 |
Imagine | +0.107 |
breathtaking | +0.106 |
Neg logit contributions | |
ried | -0.092 |
ck | -0.085 |
ases | -0.083 |
ies | -0.074 |
aunts | -0.074 |
Token | and |
Token ID | 290 |
Feature activation | -8.22e-01 |
Pos logit contributions | |
brand | +0.067 |
� | +0.065 |
truly | +0.063 |
breathtaking | +0.061 |
Imagine | +0.060 |
Neg logit contributions | |
ried | -0.044 |
ck | -0.040 |
ases | -0.038 |
ies | -0.035 |
steer | -0.034 |
Token | books |
Token ID | 3835 |
Feature activation | -3.52e+00 |
Pos logit contributions | |
brand | +0.027 |
� | +0.026 |
truly | +0.025 |
breathtaking | +0.024 |
Imagine | +0.024 |
Neg logit contributions | |
ried | -0.024 |
ck | -0.022 |
ases | -0.022 |
aunts | -0.019 |
race | -0.019 |
Token | a |
Token ID | 257 |
Feature activation | -3.80e+00 |
Pos logit contributions | |
ried | +0.172 |
ck | +0.152 |
ases | +0.149 |
aunts | +0.137 |
beh | +0.135 |
Neg logit contributions | |
brand | -0.230 |
truly | -0.216 |
� | -0.213 |
Not | -0.210 |
sky | -0.210 |
Token | new |
Token ID | 649 |
Feature activation | -2.89e+00 |
Pos logit contributions | |
� | +0.115 |
Not | +0.108 |
brand | +0.105 |
Cookies | +0.105 |
Too | +0.102 |
Neg logit contributions | |
ck | -0.100 |
ried | -0.100 |
share | -0.094 |
steer | -0.093 |
leave | -0.092 |
Token | lake |
Token ID | 13546 |
Feature activation | -9.58e+00 |
Pos logit contributions | |
brand | +0.071 |
truly | +0.063 |
sky | +0.063 |
Imagine | +0.060 |
� | +0.060 |
Neg logit contributions | |
ried | -0.114 |
ck | -0.107 |
ases | -0.105 |
ped | -0.099 |
beh | -0.099 |
Token | . |
Token ID | 13 |
Feature activation | +2.57e+00 |
Pos logit contributions | |
brand | +0.361 |
sky | +0.345 |
truly | +0.340 |
� | +0.338 |
Imagine | +0.325 |
Neg logit contributions | |
ried | -0.225 |
ck | -0.217 |
ases | -0.206 |
aunts | -0.187 |
beh | -0.184 |
Token | Dad |
Token ID | 17415 |
Feature activation | +6.25e+00 |
Pos logit contributions | |
ried | +0.082 |
ases | +0.076 |
ck | +0.073 |
aunts | +0.068 |
ucks | +0.066 |
Neg logit contributions | |
brand | -0.081 |
truly | -0.073 |
breathtaking | -0.070 |
� | -0.070 |
Imagine | -0.069 |
Token | says |
Token ID | 1139 |
Feature activation | +5.35e+00 |
Pos logit contributions | |
ried | +0.207 |
ck | +0.192 |
ases | +0.187 |
beh | +0.177 |
aunts | +0.175 |
Neg logit contributions | |
brand | -0.183 |
truly | -0.170 |
sky | -0.167 |
� | -0.164 |
Imagine | -0.163 |
Token | it |
Token ID | 340 |
Feature activation | -8.39e+00 |
Pos logit contributions | |
ried | +0.172 |
ck | +0.160 |
ases | +0.155 |
aunts | +0.143 |
race | +0.138 |
Neg logit contributions | |
brand | -0.165 |
� | -0.151 |
truly | -0.150 |
sky | -0.144 |
Imagine | -0.144 |
Token | is |
Token ID | 318 |
Feature activation | -5.45e+00 |
Pos logit contributions | |
brand | +0.211 |
� | +0.190 |
truly | +0.179 |
breathtaking | +0.175 |
Finally | +0.173 |
Neg logit contributions | |
ried | -0.325 |
ck | -0.293 |
ases | -0.288 |
aunts | -0.270 |
ped | -0.269 |
Token | very |
Token ID | 845 |
Feature activation | -1.02e+01 |
Pos logit contributions | |
brand | +0.151 |
� | +0.142 |
truly | +0.134 |
Imagine | +0.133 |
sky | +0.130 |
Neg logit contributions | |
ried | -0.191 |
ck | -0.178 |
ases | -0.171 |
aunts | -0.159 |
ped | -0.155 |
Token | deep |
Token ID | 2769 |
Feature activation | +4.18e-02 |
Pos logit contributions | |
brand | +0.223 |
truly | +0.196 |
breathtaking | +0.182 |
sky | +0.181 |
� | +0.181 |
Neg logit contributions | |
ried | -0.433 |
ases | -0.406 |
ck | -0.404 |
aunts | -0.380 |
race | -0.379 |
Token | and |
Token ID | 290 |
Feature activation | -8.49e-01 |
Pos logit contributions | |
ried | +0.001 |
ck | +0.001 |
ases | +0.001 |
aunts | +0.001 |
beh | +0.001 |
Neg logit contributions | |
brand | -0.001 |
sky | -0.001 |
truly | -0.001 |
breathtaking | -0.001 |
� | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
brand | +0.140 |
� | +0.131 |
truly | +0.128 |
Imagine | +0.126 |
breathtaking | +0.125 |
Neg logit contributions | |
ried | -0.087 |
ck | -0.077 |
ases | -0.075 |
aunts | -0.066 |
ies | -0.064 |
Token | There |
Token ID | 1318 |
Feature activation | -2.94e+00 |
Pos logit contributions | |
ried | +0.049 |
ases | +0.045 |
ck | +0.045 |
aunts | +0.042 |
race | +0.041 |
Neg logit contributions | |
brand | -0.037 |
truly | -0.032 |
breathtaking | -0.031 |
� | -0.031 |
Imagine | -0.030 |
Token | 's |
Token ID | 338 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
brand | +0.088 |
� | +0.081 |
truly | +0.076 |
Imagine | +0.076 |
breathtaking | +0.075 |
Neg logit contributions | |
ried | -0.097 |
ck | -0.090 |
ases | -0.087 |
ies | -0.080 |
leave | -0.080 |
Token | a |
Token ID | 257 |
Feature activation | -4.52e+00 |
Pos logit contributions | |
� | +0.053 |
brand | +0.053 |
Imagine | +0.048 |
Finally | +0.047 |
Very | +0.047 |
Neg logit contributions | |
ried | -0.070 |
ck | -0.066 |
ases | -0.063 |
leave | -0.061 |
ped | -0.058 |
Token | huge |
Token ID | 3236 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
Not | +0.124 |
� | +0.122 |
brand | +0.122 |
Cookies | +0.116 |
Eventually | +0.112 |
Neg logit contributions | |
ried | -0.131 |
ck | -0.131 |
steer | -0.116 |
ies | -0.115 |
leave | -0.115 |
Token | tree |
Token ID | 5509 |
Feature activation | -3.72e+00 |
Pos logit contributions | |
brand | +0.039 |
� | +0.036 |
truly | +0.036 |
Not | +0.035 |
Imagine | +0.034 |
Neg logit contributions | |
ried | -0.043 |
ck | -0.040 |
ases | -0.039 |
aunts | -0.036 |
race | -0.035 |
Token | with |
Token ID | 351 |
Feature activation | +3.59e+00 |
Pos logit contributions | |
brand | +0.137 |
� | +0.127 |
truly | +0.126 |
Not | +0.122 |
sky | +0.122 |
Neg logit contributions | |
ried | -0.097 |
ck | -0.088 |
ases | -0.085 |
aunts | -0.076 |
race | -0.073 |
Token | lots |
Token ID | 6041 |
Feature activation | -7.09e+00 |
Pos logit contributions | |
ried | +0.119 |
ck | +0.111 |
ases | +0.108 |
aunts | +0.099 |
ped | +0.097 |
Neg logit contributions | |
brand | -0.106 |
� | -0.098 |
truly | -0.096 |
Imagine | -0.093 |
Not | -0.092 |
Token | of |
Token ID | 286 |
Feature activation | +5.17e+00 |
Pos logit contributions | |
brand | +0.243 |
truly | +0.222 |
� | +0.221 |
sky | +0.218 |
breathtaking | +0.214 |
Neg logit contributions | |
ried | -0.207 |
ck | -0.188 |
ases | -0.182 |
aunts | -0.169 |
beh | -0.160 |
Token | leaves |
Token ID | 5667 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
ried | +0.166 |
ck | +0.157 |
ases | +0.152 |
aunts | +0.139 |
ped | +0.135 |
Neg logit contributions | |
brand | -0.157 |
� | -0.145 |
truly | -0.142 |
Imagine | -0.137 |
Not | -0.137 |
Token | . |
Token ID | 13 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
brand | +0.055 |
� | +0.051 |
truly | +0.051 |
sky | +0.050 |
breathtaking | +0.049 |
Neg logit contributions | |
ried | -0.036 |
ck | -0.033 |
ases | -0.032 |
aunts | -0.029 |
race | -0.027 |
Token | the |
Token ID | 262 |
Feature activation | -3.29e-02 |
Pos logit contributions | |
ried | +0.153 |
ck | +0.137 |
ases | +0.130 |
aunts | +0.116 |
beh | +0.110 |
Neg logit contributions | |
brand | -0.272 |
� | -0.254 |
truly | -0.253 |
sky | -0.248 |
breathtaking | -0.245 |
Token | beach |
Token ID | 10481 |
Feature activation | -5.35e+00 |
Pos logit contributions | |
� | +0.001 |
brand | +0.001 |
Not | +0.001 |
Imagine | +0.001 |
Too | +0.001 |
Neg logit contributions | |
ried | -0.001 |
ases | -0.001 |
ck | -0.001 |
steer | -0.001 |
aunts | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +2.76e+00 |
Pos logit contributions | |
brand | +0.213 |
sky | +0.204 |
truly | +0.199 |
� | +0.195 |
breathtaking | +0.190 |
Neg logit contributions | |
ried | -0.124 |
ck | -0.113 |
ases | -0.106 |
aunts | -0.098 |
race | -0.095 |
Token | One |
Token ID | 1881 |
Feature activation | -2.69e+00 |
Pos logit contributions | |
ried | +0.095 |
ases | +0.088 |
ck | +0.087 |
aunts | +0.083 |
race | +0.083 |
Neg logit contributions | |
brand | -0.075 |
truly | -0.066 |
breathtaking | -0.065 |
� | -0.063 |
fantastic | -0.060 |
Token | day |
Token ID | 1110 |
Feature activation | -8.15e+00 |
Pos logit contributions | |
brand | +0.069 |
� | +0.064 |
truly | +0.057 |
Not | +0.056 |
breathtaking | +0.056 |
Neg logit contributions | |
ried | -0.101 |
ck | -0.097 |
ases | -0.093 |
ie | -0.089 |
ies | -0.089 |
Token | , |
Token ID | 11 |
Feature activation | -2.18e+00 |
Pos logit contributions | |
brand | +0.206 |
� | +0.192 |
Not | +0.187 |
truly | +0.184 |
sky | +0.183 |
Neg logit contributions | |
ried | -0.317 |
ck | -0.293 |
ases | -0.288 |
aunts | -0.270 |
ucks | -0.263 |
Token | she |
Token ID | 673 |
Feature activation | +4.52e-01 |
Pos logit contributions | |
brand | +0.071 |
� | +0.065 |
truly | +0.065 |
Not | +0.063 |
sky | +0.063 |
Neg logit contributions | |
ried | -0.070 |
ases | -0.064 |
ck | -0.063 |
aunts | -0.058 |
beh | -0.056 |
Token | saw |
Token ID | 2497 |
Feature activation | +4.03e+00 |
Pos logit contributions | |
ried | +0.015 |
ck | +0.014 |
ases | +0.014 |
beh | +0.013 |
aunts | +0.013 |
Neg logit contributions | |
brand | -0.014 |
truly | -0.013 |
Imagine | -0.013 |
sky | -0.012 |
Not | -0.011 |
Token | a |
Token ID | 257 |
Feature activation | -3.96e+00 |
Pos logit contributions | |
ried | +0.131 |
ck | +0.124 |
ases | +0.120 |
aunts | +0.111 |
ped | +0.107 |
Neg logit contributions | |
brand | -0.121 |
� | -0.113 |
truly | -0.110 |
Not | -0.106 |
Imagine | -0.106 |
Token | purple |
Token ID | 14032 |
Feature activation | -2.37e+00 |
Pos logit contributions | |
� | +0.137 |
Not | +0.131 |
Too | +0.128 |
Eventually | +0.128 |
Finally | +0.126 |
Neg logit contributions | |
ck | -0.090 |
ies | -0.090 |
ried | -0.086 |
steer | -0.086 |
maneu | -0.077 |
Token | shell |
Token ID | 7582 |
Feature activation | -5.88e+00 |
Pos logit contributions | |
brand | +0.082 |
� | +0.076 |
truly | +0.076 |
Not | +0.074 |
Imagine | +0.073 |
Neg logit contributions | |
ried | -0.065 |
ck | -0.060 |
ases | -0.058 |
aunts | -0.053 |
ies | -0.052 |
Token | if |
Token ID | 611 |
Feature activation | +9.35e-01 |
Pos logit contributions | |
ried | +0.186 |
ck | +0.167 |
ases | +0.162 |
aunts | +0.146 |
race | +0.140 |
Neg logit contributions | |
brand | -0.278 |
� | -0.259 |
truly | -0.255 |
sky | -0.250 |
breathtaking | -0.249 |
Token | he |
Token ID | 339 |
Feature activation | +2.74e-01 |
Pos logit contributions | |
ried | +0.040 |
ases | +0.039 |
ck | +0.037 |
ucks | +0.036 |
aunts | +0.035 |
Neg logit contributions | |
brand | -0.019 |
truly | -0.019 |
Not | -0.018 |
sky | -0.017 |
Imagine | -0.016 |
Token | could |
Token ID | 714 |
Feature activation | -2.43e+00 |
Pos logit contributions | |
ried | +0.013 |
ases | +0.013 |
ck | +0.013 |
ucks | +0.012 |
aunts | +0.012 |
Neg logit contributions | |
truly | -0.004 |
brand | -0.004 |
Not | -0.003 |
Imagine | -0.003 |
may | -0.003 |
Token | go |
Token ID | 467 |
Feature activation | +5.79e+00 |
Pos logit contributions | |
� | +0.076 |
truly | +0.074 |
definitely | +0.069 |
Not | +0.068 |
brand | +0.066 |
Neg logit contributions | |
ried | -0.085 |
ped | -0.074 |
ases | -0.073 |
ucks | -0.072 |
beh | -0.072 |
Token | , |
Token ID | 11 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
ried | +0.177 |
ases | +0.169 |
ck | +0.162 |
aunts | +0.161 |
ped | +0.160 |
Neg logit contributions | |
brand | -0.197 |
truly | -0.183 |
� | -0.170 |
sky | -0.169 |
so | -0.162 |
Token | but |
Token ID | 475 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
brand | +0.036 |
� | +0.034 |
breathtaking | +0.032 |
truly | +0.032 |
Imagine | +0.031 |
Neg logit contributions | |
ried | -0.047 |
ck | -0.043 |
ases | -0.042 |
leave | -0.040 |
steer | -0.040 |
Token | they |
Token ID | 484 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
brand | +0.075 |
truly | +0.070 |
� | +0.068 |
sky | +0.068 |
Not | +0.067 |
Neg logit contributions | |
ried | -0.045 |
ck | -0.041 |
ases | -0.041 |
aunts | -0.036 |
ucks | -0.034 |
Token | just |
Token ID | 655 |
Feature activation | +2.01e+00 |
Pos logit contributions | |
ried | +0.032 |
ases | +0.029 |
ck | +0.029 |
aunts | +0.027 |
ucks | +0.026 |
Neg logit contributions | |
brand | -0.032 |
truly | -0.030 |
� | -0.029 |
Not | -0.028 |
sky | -0.028 |
Token | laughed |
Token ID | 13818 |
Feature activation | +3.03e+00 |
Pos logit contributions | |
ases | +0.085 |
ucks | +0.081 |
ried | +0.081 |
aunts | +0.079 |
race | +0.078 |
Neg logit contributions | |
brand | -0.038 |
truly | -0.036 |
may | -0.035 |
sky | -0.032 |
definitely | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | +3.05e+00 |
Pos logit contributions | |
ried | +0.088 |
ck | +0.082 |
ases | +0.081 |
aunts | +0.074 |
beh | +0.072 |
Neg logit contributions | |
brand | -0.109 |
truly | -0.098 |
sky | -0.094 |
Not | -0.092 |
breathtaking | -0.092 |
Token |
|
Token ID | 198 |
Feature activation | -3.20e+00 |
Pos logit contributions | |
ried | +0.115 |
ck | +0.107 |
ases | +0.106 |
aunts | +0.100 |
race | +0.097 |
Neg logit contributions | |
brand | -0.077 |
truly | -0.068 |
� | -0.066 |
breathtaking | -0.065 |
Imagine | -0.063 |
Token | small |
Token ID | 1402 |
Feature activation | -2.16e+00 |
Pos logit contributions | |
brand | +0.116 |
� | +0.112 |
truly | +0.108 |
Not | +0.106 |
breathtaking | +0.103 |
Neg logit contributions | |
ried | -0.128 |
ck | -0.124 |
ases | -0.116 |
aunts | -0.113 |
ies | -0.110 |
Token | city |
Token ID | 1748 |
Feature activation | -6.20e+00 |
Pos logit contributions | |
brand | +0.069 |
� | +0.063 |
truly | +0.063 |
Not | +0.061 |
breathtaking | +0.060 |
Neg logit contributions | |
ried | -0.066 |
ck | -0.061 |
ases | -0.060 |
aunts | -0.055 |
ies | -0.053 |
Token | . |
Token ID | 13 |
Feature activation | +2.05e+00 |
Pos logit contributions | |
brand | +0.260 |
� | +0.243 |
truly | +0.243 |
sky | +0.239 |
Imagine | +0.236 |
Neg logit contributions | |
ried | -0.132 |
ck | -0.117 |
ases | -0.112 |
aunts | -0.098 |
beh | -0.092 |
Token |
|
Token ID | 198 |
Feature activation | -4.10e+00 |
Pos logit contributions | |
ried | +0.039 |
ases | +0.035 |
her | +0.035 |
share | +0.032 |
him | +0.032 |
Neg logit contributions | |
brand | -0.085 |
truly | -0.075 |
� | -0.071 |
breathtaking | -0.071 |
Finally | -0.070 |
Token |
|
Token ID | 198 |
Feature activation | -4.18e+00 |
Pos logit contributions | |
brand | +0.116 |
� | +0.101 |
truly | +0.098 |
breathtaking | +0.095 |
fact | +0.094 |
Neg logit contributions | |
ried | -0.138 |
ck | -0.134 |
ases | -0.129 |
aunts | -0.125 |
ucks | -0.120 |
Token | One |
Token ID | 3198 |
Feature activation | -3.17e+00 |
Pos logit contributions | |
brand | +0.139 |
� | +0.127 |
may | +0.114 |
truly | +0.114 |
fact | +0.110 |
Neg logit contributions | |
ried | -0.116 |
ck | -0.113 |
aunts | -0.106 |
ases | -0.104 |
ie | -0.103 |
Token | day |
Token ID | 1110 |
Feature activation | -8.92e+00 |
Pos logit contributions | |
brand | +0.067 |
� | +0.059 |
truly | +0.057 |
Imagine | +0.053 |
breathtaking | +0.053 |
Neg logit contributions | |
ried | -0.133 |
ck | -0.126 |
ases | -0.123 |
aunts | -0.117 |
race | -0.113 |
Token | , |
Token ID | 11 |
Feature activation | -2.46e+00 |
Pos logit contributions | |
brand | +0.333 |
truly | +0.303 |
� | +0.295 |
breathtaking | +0.293 |
Imagine | +0.283 |
Neg logit contributions | |
ried | -0.231 |
ck | -0.210 |
ases | -0.206 |
ies | -0.191 |
aunts | -0.187 |
Token | she |
Token ID | 673 |
Feature activation | -2.15e-01 |
Pos logit contributions | |
brand | +0.102 |
� | +0.095 |
truly | +0.093 |
breathtaking | +0.091 |
Imagine | +0.089 |
Neg logit contributions | |
ried | -0.052 |
ck | -0.044 |
ases | -0.041 |
ies | -0.040 |
ie | -0.039 |
Token | decided |
Token ID | 3066 |
Feature activation | +2.88e+00 |
Pos logit contributions | |
Imagine | +0.006 |
brand | +0.006 |
sky | +0.006 |
truly | +0.006 |
Not | +0.005 |
Neg logit contributions | |
ried | -0.007 |
ck | -0.007 |
ases | -0.007 |
aunts | -0.006 |
beh | -0.006 |
Token | to |
Token ID | 284 |
Feature activation | +5.59e+00 |
Pos logit contributions | |
ried | +0.080 |
ases | +0.074 |
ck | +0.074 |
aunts | +0.066 |
ucks | +0.063 |
Neg logit contributions | |
brand | -0.104 |
truly | -0.096 |
sky | -0.094 |
Not | -0.093 |
Imagine | -0.093 |
Token | how |
Token ID | 703 |
Feature activation | -1.01e+01 |
Pos logit contributions | |
ried | +0.126 |
ck | +0.116 |
ases | +0.114 |
aunts | +0.104 |
race | +0.100 |
Neg logit contributions | |
brand | -0.130 |
� | -0.120 |
truly | -0.118 |
Imagine | -0.115 |
sky | -0.115 |
Token | to |
Token ID | 284 |
Feature activation | +5.55e+00 |
Pos logit contributions | |
brand | +0.342 |
truly | +0.313 |
� | +0.311 |
sky | +0.302 |
Not | +0.302 |
Neg logit contributions | |
ried | -0.298 |
ck | -0.275 |
ases | -0.266 |
aunts | -0.245 |
race | -0.233 |
Token | improve |
Token ID | 2987 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
ried | +0.169 |
ck | +0.145 |
ases | +0.144 |
aunts | +0.134 |
ucks | +0.134 |
Neg logit contributions | |
brand | -0.185 |
� | -0.180 |
truly | -0.179 |
Not | -0.171 |
sky | -0.167 |
Token | ." |
Token ID | 526 |
Feature activation | +9.64e-01 |
Pos logit contributions | |
ried | +0.035 |
ck | +0.032 |
ases | +0.031 |
aunts | +0.028 |
race | +0.027 |
Neg logit contributions | |
brand | -0.053 |
� | -0.049 |
truly | -0.048 |
sky | -0.047 |
Imagine | -0.047 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +7.57e-01 |
Pos logit contributions | |
ried | +0.035 |
ck | +0.032 |
ases | +0.032 |
aunts | +0.030 |
ped | +0.030 |
Neg logit contributions | |
brand | -0.025 |
truly | -0.022 |
breathtaking | -0.022 |
� | -0.021 |
Imagine | -0.021 |
Token | There |
Token ID | 1858 |
Feature activation | -3.37e+00 |
Pos logit contributions | |
ried | +0.032 |
ck | +0.030 |
ases | +0.029 |
aunts | +0.028 |
beh | +0.027 |
Neg logit contributions | |
brand | -0.016 |
� | -0.014 |
truly | -0.014 |
sky | -0.013 |
Imagine | -0.013 |
Token | was |
Token ID | 373 |
Feature activation | -3.66e+00 |
Pos logit contributions | |
brand | +0.072 |
� | +0.065 |
truly | +0.064 |
sky | +0.062 |
Imagine | +0.060 |
Neg logit contributions | |
ried | -0.140 |
ck | -0.132 |
ases | -0.129 |
aunts | -0.122 |
beh | -0.119 |
Token | a |
Token ID | 257 |
Feature activation | -3.84e+00 |
Pos logit contributions | |
brand | +0.127 |
� | +0.120 |
may | +0.113 |
daunting | +0.112 |
truly | +0.112 |
Neg logit contributions | |
ried | -0.089 |
ck | -0.087 |
ases | -0.087 |
ped | -0.080 |
his | -0.078 |
Token | cat |
Token ID | 3797 |
Feature activation | +1.90e+00 |
Pos logit contributions | |
brand | +0.107 |
� | +0.097 |
truly | +0.096 |
sky | +0.092 |
Not | +0.092 |
Neg logit contributions | |
ried | -0.134 |
ck | -0.125 |
ases | -0.122 |
aunts | -0.114 |
race | -0.110 |
Token | . |
Token ID | 13 |
Feature activation | +2.36e+00 |
Pos logit contributions | |
ried | +0.069 |
ases | +0.062 |
ck | +0.061 |
ucks | +0.058 |
aunts | +0.056 |
Neg logit contributions | |
brand | -0.055 |
truly | -0.053 |
sky | -0.053 |
� | -0.050 |
Imagine | -0.049 |
Token | She |
Token ID | 1375 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
ried | +0.075 |
ck | +0.069 |
ases | +0.067 |
aunts | +0.062 |
race | +0.059 |
Neg logit contributions | |
brand | -0.075 |
� | -0.068 |
truly | -0.068 |
sky | -0.066 |
Not | -0.065 |
Token | . |
Token ID | 13 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
ried | +0.144 |
beh | +0.134 |
ases | +0.134 |
ck | +0.134 |
aunts | +0.130 |
Neg logit contributions | |
brand | -0.129 |
� | -0.126 |
so | -0.123 |
much | -0.121 |
Too | -0.119 |
Token |
|
Token ID | 198 |
Feature activation | -4.28e+00 |
Pos logit contributions | |
ried | +0.065 |
ases | +0.061 |
ucks | +0.058 |
aunts | +0.058 |
pered | +0.056 |
Neg logit contributions | |
brand | -0.058 |
truly | -0.053 |
breathtaking | -0.051 |
fantastic | -0.048 |
� | -0.046 |
Token |
|
Token ID | 198 |
Feature activation | -4.29e+00 |
Pos logit contributions | |
brand | +0.096 |
� | +0.084 |
truly | +0.083 |
breathtaking | +0.078 |
sky | +0.078 |
Neg logit contributions | |
ried | -0.174 |
ck | -0.165 |
ases | -0.162 |
aunts | -0.152 |
race | -0.147 |
Token | " |
Token ID | 1 |
Feature activation | -2.34e-01 |
Pos logit contributions | |
brand | +0.138 |
� | +0.129 |
sky | +0.129 |
truly | +0.128 |
Imagine | +0.127 |
Neg logit contributions | |
ried | -0.129 |
ck | -0.115 |
ases | -0.113 |
beh | -0.103 |
race | -0.102 |
Token | Me |
Token ID | 5308 |
Feature activation | -2.02e+00 |
Pos logit contributions | |
brand | +0.008 |
truly | +0.007 |
impossible | +0.007 |
� | +0.007 |
case | +0.007 |
Neg logit contributions | |
ried | -0.006 |
ck | -0.006 |
ases | -0.006 |
aunts | -0.006 |
steer | -0.006 |
Token | too |
Token ID | 1165 |
Feature activation | -1.16e+01 |
Pos logit contributions | |
brand | +0.052 |
truly | +0.044 |
� | +0.044 |
breathtaking | +0.043 |
Imagine | +0.043 |
Neg logit contributions | |
ried | -0.078 |
ck | -0.073 |
ases | -0.072 |
aunts | -0.067 |
beh | -0.064 |
Token | , |
Token ID | 11 |
Feature activation | -2.08e+00 |
Pos logit contributions | |
brand | +0.352 |
� | +0.323 |
truly | +0.320 |
sky | +0.313 |
Not | +0.310 |
Neg logit contributions | |
ried | -0.383 |
ck | -0.352 |
ases | -0.344 |
aunts | -0.319 |
race | -0.316 |
Token | me |
Token ID | 502 |
Feature activation | +4.27e+00 |
Pos logit contributions | |
brand | +0.088 |
� | +0.082 |
truly | +0.082 |
sky | +0.080 |
Imagine | +0.080 |
Neg logit contributions | |
ried | -0.044 |
ck | -0.039 |
ases | -0.037 |
aunts | -0.032 |
race | -0.030 |
Token | too |
Token ID | 1165 |
Feature activation | -1.16e+01 |
Pos logit contributions | |
ried | +0.138 |
ck | +0.128 |
ases | +0.125 |
aunts | +0.118 |
beh | +0.114 |
Neg logit contributions | |
brand | -0.129 |
� | -0.119 |
truly | -0.118 |
sky | -0.113 |
Not | -0.113 |
Token | !" |
Token ID | 2474 |
Feature activation | +7.63e-01 |
Pos logit contributions | |
brand | +0.353 |
� | +0.327 |
truly | +0.327 |
sky | +0.322 |
Not | +0.319 |
Neg logit contributions | |
ried | -0.383 |
ases | -0.342 |
ck | -0.340 |
race | -0.314 |
beh | -0.313 |
Token | Ben |
Token ID | 3932 |
Feature activation | +9.26e+00 |
Pos logit contributions | |
ried | +0.017 |
ck | +0.014 |
ases | +0.013 |
beh | +0.012 |
race | +0.012 |
Neg logit contributions | |
brand | -0.032 |
� | -0.031 |
sky | -0.030 |
Not | -0.030 |
truly | -0.030 |
Token | was |
Token ID | 373 |
Feature activation | -3.87e+00 |
Pos logit contributions | |
brand | +0.120 |
daunting | +0.105 |
� | +0.104 |
breathtaking | +0.103 |
Finally | +0.102 |
Neg logit contributions | |
ried | -0.082 |
him | -0.080 |
his | -0.079 |
ie | -0.078 |
ck | -0.076 |
Token | an |
Token ID | 281 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
brand | +0.139 |
� | +0.128 |
truly | +0.125 |
daunting | +0.125 |
may | +0.123 |
Neg logit contributions | |
ck | -0.091 |
ried | -0.089 |
ases | -0.086 |
him | -0.078 |
ped | -0.078 |
Token | attractive |
Token ID | 10966 |
Feature activation | -3.12e+00 |
Pos logit contributions | |
brand | +0.041 |
truly | +0.036 |
sky | +0.035 |
breathtaking | +0.034 |
Not | +0.033 |
Neg logit contributions | |
ried | -0.062 |
ases | -0.056 |
ck | -0.056 |
aunts | -0.052 |
race | -0.051 |
Token | valley |
Token ID | 19272 |
Feature activation | -7.76e+00 |
Pos logit contributions | |
brand | +0.101 |
sky | +0.093 |
� | +0.092 |
truly | +0.090 |
breathtaking | +0.086 |
Neg logit contributions | |
ried | -0.104 |
ck | -0.094 |
ases | -0.093 |
beh | -0.085 |
aunts | -0.083 |
Token | . |
Token ID | 13 |
Feature activation | +2.17e+00 |
Pos logit contributions | |
brand | +0.293 |
sky | +0.281 |
truly | +0.278 |
� | +0.274 |
breathtaking | +0.263 |
Neg logit contributions | |
ried | -0.193 |
ck | -0.186 |
ases | -0.178 |
aunts | -0.155 |
ucks | -0.154 |
Token | It |
Token ID | 632 |
Feature activation | -1.01e+01 |
Pos logit contributions | |
ried | +0.077 |
ases | +0.072 |
ck | +0.071 |
race | +0.068 |
aunts | +0.067 |
Neg logit contributions | |
brand | -0.058 |
truly | -0.051 |
� | -0.048 |
breathtaking | -0.047 |
Imagine | -0.046 |
Token | was |
Token ID | 373 |
Feature activation | -3.89e+00 |
Pos logit contributions | |
brand | +0.322 |
� | +0.301 |
breathtaking | +0.268 |
Not | +0.266 |
Finally | +0.264 |
Neg logit contributions | |
ried | -0.335 |
ck | -0.290 |
ran | -0.277 |
ie | -0.271 |
ases | -0.268 |
Token | a |
Token ID | 257 |
Feature activation | -3.99e+00 |
Pos logit contributions | |
� | +0.122 |
brand | +0.109 |
Not | +0.101 |
Very | +0.101 |
Finally | +0.100 |
Neg logit contributions | |
ried | -0.121 |
ck | -0.109 |
his | -0.108 |
her | -0.102 |
ases | -0.101 |
Token | lovely |
Token ID | 14081 |
Feature activation | -4.02e+00 |
Pos logit contributions | |
� | +0.117 |
Not | +0.112 |
Eventually | +0.110 |
pill | +0.110 |
brand | +0.110 |
Neg logit contributions | |
share | -0.097 |
ried | -0.095 |
ck | -0.092 |
ies | -0.092 |
pet | -0.092 |
Token | place |
Token ID | 1295 |
Feature activation | -8.31e+00 |
Pos logit contributions | |
brand | +0.108 |
� | +0.105 |
Not | +0.102 |
Finally | +0.101 |
Eventually | +0.099 |
Neg logit contributions | |
ried | -0.128 |
ie | -0.123 |
ies | -0.121 |
ck | -0.115 |
leave | -0.113 |
Token | - |
Token ID | 532 |
Feature activation | -2.99e+00 |
Pos logit contributions | |
brand | +0.310 |
� | +0.300 |
Imagine | +0.287 |
Not | +0.285 |
Finally | +0.284 |
Neg logit contributions | |
ried | -0.206 |
ck | -0.178 |
ases | -0.173 |
ies | -0.159 |
aunts | -0.158 |
Token | to |
Token ID | 284 |
Feature activation | +6.01e+00 |
Pos logit contributions | |
ried | +0.076 |
ck | +0.068 |
ases | +0.065 |
aunts | +0.058 |
race | +0.057 |
Neg logit contributions | |
brand | -0.130 |
� | -0.124 |
truly | -0.121 |
Imagine | -0.118 |
breathtaking | -0.118 |
Token | heal |
Token ID | 12035 |
Feature activation | +2.60e+00 |
Pos logit contributions | |
ried | +0.191 |
ck | +0.158 |
ases | +0.152 |
rying | +0.146 |
ped | +0.145 |
Neg logit contributions | |
brand | -0.208 |
truly | -0.199 |
Not | -0.197 |
� | -0.197 |
sky | -0.190 |
Token | . |
Token ID | 13 |
Feature activation | +2.32e+00 |
Pos logit contributions | |
ried | +0.060 |
ases | +0.050 |
ck | +0.049 |
aunts | +0.044 |
race | +0.043 |
Neg logit contributions | |
brand | -0.108 |
truly | -0.100 |
� | -0.098 |
sky | -0.097 |
Not | -0.097 |
Token | He |
Token ID | 679 |
Feature activation | +2.21e-01 |
Pos logit contributions | |
ried | +0.060 |
ases | +0.053 |
ck | +0.049 |
aunts | +0.049 |
race | +0.048 |
Neg logit contributions | |
brand | -0.086 |
truly | -0.079 |
� | -0.078 |
breathtaking | -0.077 |
Imagine | -0.075 |
Token | was |
Token ID | 373 |
Feature activation | -3.59e+00 |
Pos logit contributions | |
ases | +0.008 |
ried | +0.008 |
ucks | +0.007 |
ck | +0.007 |
aunts | +0.007 |
Neg logit contributions | |
brand | -0.006 |
truly | -0.006 |
Imagine | -0.006 |
sky | -0.005 |
Not | -0.005 |
Token | so |
Token ID | 523 |
Feature activation | -1.15e+01 |
Pos logit contributions | |
brand | +0.073 |
� | +0.068 |
truly | +0.062 |
breathtaking | +0.061 |
sky | +0.061 |
Neg logit contributions | |
ried | -0.152 |
ck | -0.144 |
ases | -0.138 |
aunts | -0.131 |
beh | -0.129 |
Token | eager |
Token ID | 11069 |
Feature activation | +3.10e-01 |
Pos logit contributions | |
brand | +0.319 |
� | +0.289 |
truly | +0.285 |
sky | +0.276 |
breathtaking | +0.275 |
Neg logit contributions | |
ried | -0.407 |
ck | -0.383 |
ases | -0.370 |
aunts | -0.347 |
beh | -0.335 |
Token | to |
Token ID | 284 |
Feature activation | +6.07e+00 |
Pos logit contributions | |
ried | +0.005 |
ck | +0.005 |
ases | +0.004 |
ies | +0.004 |
race | +0.004 |
Neg logit contributions | |
brand | -0.014 |
� | -0.013 |
truly | -0.013 |
Finally | -0.013 |
breathtaking | -0.013 |
Token | help |
Token ID | 1037 |
Feature activation | +4.82e+00 |
Pos logit contributions | |
ried | +0.166 |
ases | +0.142 |
ck | +0.139 |
aunts | +0.127 |
beh | +0.123 |
Neg logit contributions | |
brand | -0.227 |
truly | -0.212 |
� | -0.210 |
Not | -0.207 |
sky | -0.204 |
Token | people |
Token ID | 661 |
Feature activation | +2.03e+00 |
Pos logit contributions | |
ried | +0.124 |
ases | +0.098 |
ck | +0.093 |
aunts | +0.083 |
beh | +0.083 |
Neg logit contributions | |
brand | -0.196 |
truly | -0.186 |
Not | -0.182 |
sky | -0.179 |
lot | -0.172 |
Token | and |
Token ID | 290 |
Feature activation | -7.00e-01 |
Pos logit contributions | |
ried | +0.050 |
ases | +0.045 |
ck | +0.044 |
aunts | +0.040 |
ucks | +0.038 |
Neg logit contributions | |
brand | -0.079 |
truly | -0.074 |
� | -0.072 |
Not | -0.071 |
sky | -0.070 |
Token | |
Token ID | 220 |
Feature activation | -1.22e+01 |
Pos logit contributions | |
brand | +0.095 |
� | +0.083 |
truly | +0.082 |
sky | +0.078 |
Imagine | +0.077 |
Neg logit contributions | |
ried | -0.188 |
ck | -0.177 |
ases | -0.174 |
aunts | -0.164 |
race | -0.159 |
Token |
|
Token ID | 198 |
Feature activation | -4.62e+00 |
Pos logit contributions | |
brand | +0.241 |
� | +0.213 |
truly | +0.208 |
Not | +0.201 |
sky | +0.198 |
Neg logit contributions | |
ried | -0.526 |
ck | -0.502 |
ases | -0.487 |
aunts | -0.463 |
beh | -0.452 |
Token | The |
Token ID | 464 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
brand | +0.120 |
truly | +0.109 |
� | +0.108 |
Imagine | +0.104 |
Not | +0.104 |
Neg logit contributions | |
ried | -0.172 |
ck | -0.160 |
ases | -0.157 |
aunts | -0.146 |
race | -0.141 |
Token | youth |
Token ID | 6205 |
Feature activation | +3.83e+00 |
Pos logit contributions | |
brand | +0.042 |
� | +0.038 |
truly | +0.038 |
Not | +0.037 |
Imagine | +0.037 |
Neg logit contributions | |
ried | -0.043 |
ck | -0.040 |
ases | -0.039 |
aunts | -0.036 |
race | -0.035 |
Token | was |
Token ID | 373 |
Feature activation | -4.07e+00 |
Pos logit contributions | |
ried | +0.149 |
ases | +0.143 |
ck | +0.139 |
aunts | +0.132 |
ucks | +0.131 |
Neg logit contributions | |
brand | -0.087 |
truly | -0.084 |
sky | -0.079 |
� | -0.079 |
Imagine | -0.077 |
Token | so |
Token ID | 523 |
Feature activation | -1.21e+01 |
Pos logit contributions | |
brand | +0.080 |
� | +0.077 |
Not | +0.068 |
Imagine | +0.067 |
sky | +0.066 |
Neg logit contributions | |
ried | -0.173 |
ck | -0.167 |
ases | -0.156 |
ped | -0.148 |
aunts | -0.148 |
Token | sad |
Token ID | 6507 |
Feature activation | -5.03e+00 |
Pos logit contributions | |
brand | +0.330 |
� | +0.306 |
truly | +0.287 |
sky | +0.285 |
Imagine | +0.283 |
Neg logit contributions | |
ried | -0.432 |
ck | -0.411 |
ases | -0.388 |
aunts | -0.365 |
beh | -0.356 |
Token | that |
Token ID | 326 |
Feature activation | -3.66e+00 |
Pos logit contributions | |
brand | +0.155 |
� | +0.141 |
truly | +0.141 |
sky | +0.136 |
Not | +0.135 |
Neg logit contributions | |
ried | -0.163 |
ck | -0.151 |
ases | -0.147 |
aunts | -0.136 |
race | -0.131 |
Token | the |
Token ID | 262 |
Feature activation | -7.36e-01 |
Pos logit contributions | |
brand | +0.103 |
truly | +0.097 |
sky | +0.094 |
Not | +0.091 |
� | +0.090 |
Neg logit contributions | |
ried | -0.131 |
ck | -0.121 |
ases | -0.121 |
aunts | -0.111 |
ucks | -0.110 |
Token | balloon |
Token ID | 21190 |
Feature activation | -4.19e+00 |
Pos logit contributions | |
brand | +0.023 |
� | +0.021 |
truly | +0.021 |
sky | +0.020 |
breathtaking | +0.020 |
Neg logit contributions | |
ried | -0.024 |
ck | -0.022 |
ases | -0.021 |
aunts | -0.020 |
race | -0.019 |
Token | was |
Token ID | 373 |
Feature activation | -4.10e+00 |
Pos logit contributions | |
brand | +0.105 |
� | +0.093 |
truly | +0.093 |
sky | +0.090 |
Imagine | +0.088 |
Neg logit contributions | |
ried | -0.159 |
ck | -0.149 |
ases | -0.147 |
aunts | -0.137 |
race | -0.133 |
Token | not |
Token ID | 407 |
Feature activation | -2.37e+00 |
Pos logit contributions | |
brand | +0.159 |
truly | +0.142 |
� | +0.136 |
sky | +0.134 |
breathtaking | +0.133 |
Neg logit contributions | |
ried | -0.226 |
ck | -0.209 |
ases | -0.208 |
aunts | -0.196 |
race | -0.189 |
Token | available |
Token ID | 1695 |
Feature activation | -2.60e+00 |
Pos logit contributions | |
brand | +0.064 |
truly | +0.060 |
Not | +0.055 |
sky | +0.055 |
breathtaking | +0.054 |
Neg logit contributions | |
ried | -0.085 |
ases | -0.080 |
ck | -0.078 |
aunts | -0.076 |
ucks | -0.073 |
Token | for |
Token ID | 329 |
Feature activation | +4.64e+00 |
Pos logit contributions | |
brand | +0.095 |
truly | +0.089 |
sky | +0.087 |
Not | +0.086 |
impossible | +0.083 |
Neg logit contributions | |
ried | -0.068 |
ck | -0.065 |
ases | -0.060 |
aunts | -0.059 |
ucks | -0.055 |
Token | you |
Token ID | 345 |
Feature activation | +2.69e+00 |
Pos logit contributions | |
ried | +0.141 |
ck | +0.120 |
ases | +0.119 |
aunts | +0.116 |
ucks | +0.109 |
Neg logit contributions | |
brand | -0.156 |
truly | -0.144 |
Not | -0.143 |
sky | -0.142 |
� | -0.141 |
Token | . |
Token ID | 13 |
Feature activation | +1.71e+00 |
Pos logit contributions | |
ried | +0.060 |
ck | +0.055 |
ases | +0.052 |
aunts | +0.048 |
ucks | +0.045 |
Neg logit contributions | |
brand | -0.108 |
truly | -0.102 |
Not | -0.100 |
� | -0.099 |
sky | -0.099 |
Token | It |
Token ID | 632 |
Feature activation | -1.01e+01 |
Pos logit contributions | |
ried | +0.067 |
ck | +0.063 |
ases | +0.061 |
aunts | +0.058 |
race | +0.056 |
Neg logit contributions | |
brand | -0.041 |
� | -0.036 |
truly | -0.036 |
sky | -0.035 |
Not | -0.034 |
Token | is |
Token ID | 318 |
Feature activation | -6.01e+00 |
Pos logit contributions | |
brand | +0.297 |
� | +0.275 |
breathtaking | +0.254 |
Imagine | +0.250 |
Finally | +0.247 |
Neg logit contributions | |
ried | -0.352 |
ck | -0.331 |
ases | -0.312 |
leave | -0.288 |
aunts | -0.287 |
Token | too |
Token ID | 1165 |
Feature activation | -1.15e+01 |
Pos logit contributions | |
brand | +0.165 |
� | +0.159 |
Imagine | +0.147 |
truly | +0.145 |
sky | +0.143 |
Neg logit contributions | |
ried | -0.210 |
ck | -0.199 |
ases | -0.188 |
aunts | -0.173 |
ped | -0.170 |
Token | fast |
Token ID | 3049 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
brand | +0.324 |
� | +0.313 |
truly | +0.289 |
Imagine | +0.287 |
Not | +0.281 |
Neg logit contributions | |
ried | -0.391 |
ck | -0.377 |
ases | -0.352 |
aunts | -0.330 |
steer | -0.326 |
Token | and |
Token ID | 290 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
ried | +0.039 |
ck | +0.037 |
ases | +0.035 |
aunts | +0.033 |
beh | +0.032 |
Neg logit contributions | |
brand | -0.044 |
sky | -0.039 |
truly | -0.039 |
� | -0.039 |
Not | -0.038 |
Token | too |
Token ID | 1165 |
Feature activation | -1.15e+01 |
Pos logit contributions | |
brand | +0.035 |
� | +0.033 |
truly | +0.031 |
Imagine | +0.030 |
sky | +0.030 |
Neg logit contributions | |
ried | -0.045 |
ck | -0.042 |
ases | -0.040 |
aunts | -0.037 |
steer | -0.037 |